diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5038 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 8.0,
+  "eval_steps": 500,
+  "global_step": 1431,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.011180992313067784,
+      "grad_norm": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 1.2878,
+      "step": 2
+    },
+    {
+      "epoch": 0.02236198462613557,
+      "grad_norm": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 1.392,
+      "step": 4
+    },
+    {
+      "epoch": 0.033542976939203356,
+      "grad_norm": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 1.3594,
+      "step": 6
+    },
+    {
+      "epoch": 0.04472396925227114,
+      "grad_norm": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 1.2958,
+      "step": 8
+    },
+    {
+      "epoch": 0.055904961565338925,
+      "grad_norm": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 1.3475,
+      "step": 10
+    },
+    {
+      "epoch": 0.06708595387840671,
+      "grad_norm": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 1.2303,
+      "step": 12
+    },
+    {
+      "epoch": 0.07826694619147449,
+      "grad_norm": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 1.1964,
+      "step": 14
+    },
+    {
+      "epoch": 0.08944793850454227,
+      "grad_norm": 0.0,
+      "learning_rate": 5e-06,
+      "loss": 1.3328,
+      "step": 16
+    },
+    {
+      "epoch": 0.10062893081761007,
+      "grad_norm": 6.944002672340658,
+      "learning_rate": 4.999996106235862e-06,
+      "loss": 1.3134,
+      "step": 18
+    },
+    {
+      "epoch": 0.11180992313067785,
+      "grad_norm": 7.800497498064014,
+      "learning_rate": 4.999964956195521e-06,
+      "loss": 1.1147,
+      "step": 20
+    },
+    {
+      "epoch": 0.12299091544374563,
+      "grad_norm": 4.4662495771497355,
+      "learning_rate": 4.999902656502973e-06,
+      "loss": 1.025,
+      "step": 22
+    },
+    {
+      "epoch": 0.13417190775681342,
+      "grad_norm": 4.026851738528776,
+      "learning_rate": 4.999809207934472e-06,
+      "loss": 1.0448,
+      "step": 24
+    },
+    {
+      "epoch": 0.1453529000698812,
+      "grad_norm": 5.658278761851693,
+      "learning_rate": 4.999684611654392e-06,
+      "loss": 0.9826,
+      "step": 26
+    },
+    {
+      "epoch": 0.15653389238294899,
+      "grad_norm": 3.9275877006609505,
+      "learning_rate": 4.9995288692152046e-06,
+      "loss": 0.9627,
+      "step": 28
+    },
+    {
+      "epoch": 0.16771488469601678,
+      "grad_norm": 3.634771950296262,
+      "learning_rate": 4.9993419825574686e-06,
+      "loss": 0.9476,
+      "step": 30
+    },
+    {
+      "epoch": 0.17889587700908455,
+      "grad_norm": 4.604406424526374,
+      "learning_rate": 4.9992368608591775e-06,
+      "loss": 0.9414,
+      "step": 32
+    },
+    {
+      "epoch": 0.19007686932215234,
+      "grad_norm": 5.708200502114745,
+      "learning_rate": 4.999003262361029e-06,
+      "loss": 0.9572,
+      "step": 34
+    },
+    {
+      "epoch": 0.20125786163522014,
+      "grad_norm": 5.020134712294459,
+      "learning_rate": 4.998738526193412e-06,
+      "loss": 0.9544,
+      "step": 36
+    },
+    {
+      "epoch": 0.2124388539482879,
+      "grad_norm": 4.643332496496484,
+      "learning_rate": 4.998442655654946e-06,
+      "loss": 0.8504,
+      "step": 38
+    },
+    {
+      "epoch": 0.2236198462613557,
+      "grad_norm": 4.7843514072232125,
+      "learning_rate": 4.998115654432191e-06,
+      "loss": 0.914,
+      "step": 40
+    },
+    {
+      "epoch": 0.2348008385744235,
+      "grad_norm": 3.973113705087721,
+      "learning_rate": 4.997757526599592e-06,
+      "loss": 0.8303,
+      "step": 42
+    },
+    {
+      "epoch": 0.24598183088749126,
+      "grad_norm": 5.753323652117126,
+      "learning_rate": 4.9973682766194355e-06,
+      "loss": 0.8916,
+      "step": 44
+    },
+    {
+      "epoch": 0.25716282320055905,
+      "grad_norm": 4.00607759948128,
+      "learning_rate": 4.996947909341789e-06,
+      "loss": 0.9391,
+      "step": 46
+    },
+    {
+      "epoch": 0.26834381551362685,
+      "grad_norm": 4.73751358896988,
+      "learning_rate": 4.996496430004446e-06,
+      "loss": 0.8445,
+      "step": 48
+    },
+    {
+      "epoch": 0.27952480782669464,
+      "grad_norm": 3.801634673248135,
+      "learning_rate": 4.9960138442328535e-06,
+      "loss": 0.8354,
+      "step": 50
+    },
+    {
+      "epoch": 0.2907058001397624,
+      "grad_norm": 4.998706656181077,
+      "learning_rate": 4.9955001580400475e-06,
+      "loss": 0.8556,
+      "step": 52
+    },
+    {
+      "epoch": 0.3018867924528302,
+      "grad_norm": 5.235396952388322,
+      "learning_rate": 4.994955377826577e-06,
+      "loss": 0.8821,
+      "step": 54
+    },
+    {
+      "epoch": 0.31306778476589797,
+      "grad_norm": 4.593843550283633,
+      "learning_rate": 4.994379510380421e-06,
+      "loss": 0.7965,
+      "step": 56
+    },
+    {
+      "epoch": 0.32424877707896577,
+      "grad_norm": 4.636040406542864,
+      "learning_rate": 4.993772562876909e-06,
+      "loss": 0.8576,
+      "step": 58
+    },
+    {
+      "epoch": 0.33542976939203356,
+      "grad_norm": 4.422458900120915,
+      "learning_rate": 4.993134542878631e-06,
+      "loss": 0.8388,
+      "step": 60
+    },
+    {
+      "epoch": 0.3466107617051013,
+      "grad_norm": 4.88515796654498,
+      "learning_rate": 4.992465458335335e-06,
+      "loss": 0.8427,
+      "step": 62
+    },
+    {
+      "epoch": 0.3577917540181691,
+      "grad_norm": 4.620642626620232,
+      "learning_rate": 4.991765317583841e-06,
+      "loss": 0.8088,
+      "step": 64
+    },
+    {
+      "epoch": 0.3689727463312369,
+      "grad_norm": 3.0164501013815146,
+      "learning_rate": 4.991034129347927e-06,
+      "loss": 0.7643,
+      "step": 66
+    },
+    {
+      "epoch": 0.3801537386443047,
+      "grad_norm": 4.0807085306410915,
+      "learning_rate": 4.990271902738223e-06,
+      "loss": 0.8304,
+      "step": 68
+    },
+    {
+      "epoch": 0.3913347309573725,
+      "grad_norm": 4.913983348963418,
+      "learning_rate": 4.989478647252101e-06,
+      "loss": 0.8694,
+      "step": 70
+    },
+    {
+      "epoch": 0.4025157232704403,
+      "grad_norm": 5.427166275548586,
+      "learning_rate": 4.988654372773552e-06,
+      "loss": 0.8031,
+      "step": 72
+    },
+    {
+      "epoch": 0.413696715583508,
+      "grad_norm": 4.976699288607289,
+      "learning_rate": 4.987799089573066e-06,
+      "loss": 0.7548,
+      "step": 74
+    },
+    {
+      "epoch": 0.4248777078965758,
+      "grad_norm": 5.035712861337141,
+      "learning_rate": 4.986912808307502e-06,
+      "loss": 0.7769,
+      "step": 76
+    },
+    {
+      "epoch": 0.4360587002096436,
+      "grad_norm": 5.703104314189732,
+      "learning_rate": 4.985995540019956e-06,
+      "loss": 0.7744,
+      "step": 78
+    },
+    {
+      "epoch": 0.4472396925227114,
+      "grad_norm": 3.6174332203212938,
+      "learning_rate": 4.985047296139622e-06,
+      "loss": 0.7215,
+      "step": 80
+    },
+    {
+      "epoch": 0.4584206848357792,
+      "grad_norm": 5.084461038739496,
+      "learning_rate": 4.984068088481654e-06,
+      "loss": 0.7462,
+      "step": 82
+    },
+    {
+      "epoch": 0.469601677148847,
+      "grad_norm": 5.500722673783384,
+      "learning_rate": 4.983057929247014e-06,
+      "loss": 0.7937,
+      "step": 84
+    },
+    {
+      "epoch": 0.4807826694619147,
+      "grad_norm": 5.76928743736382,
+      "learning_rate": 4.9820168310223215e-06,
+      "loss": 0.7701,
+      "step": 86
+    },
+    {
+      "epoch": 0.4919636617749825,
+      "grad_norm": 4.3638410984754366,
+      "learning_rate": 4.980944806779698e-06,
+      "loss": 0.7063,
+      "step": 88
+    },
+    {
+      "epoch": 0.5031446540880503,
+      "grad_norm": 6.6022312070502664,
+      "learning_rate": 4.979841869876603e-06,
+      "loss": 0.7829,
+      "step": 90
+    },
+    {
+      "epoch": 0.5143256464011181,
+      "grad_norm": 5.114853414480892,
+      "learning_rate": 4.97870803405567e-06,
+      "loss": 0.7419,
+      "step": 92
+    },
+    {
+      "epoch": 0.5255066387141859,
+      "grad_norm": 5.450293615821356,
+      "learning_rate": 4.977543313444534e-06,
+      "loss": 0.7428,
+      "step": 94
+    },
+    {
+      "epoch": 0.5366876310272537,
+      "grad_norm": 3.888671786201343,
+      "learning_rate": 4.976347722555655e-06,
+      "loss": 0.763,
+      "step": 96
+    },
+    {
+      "epoch": 0.5478686233403215,
+      "grad_norm": 5.580018062591517,
+      "learning_rate": 4.975121276286136e-06,
+      "loss": 0.7451,
+      "step": 98
+    },
+    {
+      "epoch": 0.5590496156533893,
+      "grad_norm": 5.244409209125885,
+      "learning_rate": 4.973863989917545e-06,
+      "loss": 0.6658,
+      "step": 100
+    },
+    {
+      "epoch": 0.570230607966457,
+      "grad_norm": 6.341201782490113,
+      "learning_rate": 4.9725758791157105e-06,
+      "loss": 0.7042,
+      "step": 102
+    },
+    {
+      "epoch": 0.5814116002795248,
+      "grad_norm": 3.63864440598579,
+      "learning_rate": 4.9712569599305415e-06,
+      "loss": 0.6859,
+      "step": 104
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 5.643540415249962,
+      "learning_rate": 4.9699072487958185e-06,
+      "loss": 0.7072,
+      "step": 106
+    },
+    {
+      "epoch": 0.6037735849056604,
+      "grad_norm": 4.518214836889502,
+      "learning_rate": 4.968526762528988e-06,
+      "loss": 0.6989,
+      "step": 108
+    },
+    {
+      "epoch": 0.6149545772187281,
+      "grad_norm": 4.813780988459217,
+      "learning_rate": 4.96711551833096e-06,
+      "loss": 0.6213,
+      "step": 110
+    },
+    {
+      "epoch": 0.6261355695317959,
+      "grad_norm": 6.534716960952802,
+      "learning_rate": 4.965673533785887e-06,
+      "loss": 0.6603,
+      "step": 112
+    },
+    {
+      "epoch": 0.6373165618448637,
+      "grad_norm": 4.694700268634709,
+      "learning_rate": 4.9642008268609455e-06,
+      "loss": 0.6458,
+      "step": 114
+    },
+    {
+      "epoch": 0.6484975541579315,
+      "grad_norm": 3.797163997052886,
+      "learning_rate": 4.962697415906118e-06,
+      "loss": 0.6208,
+      "step": 116
+    },
+    {
+      "epoch": 0.6596785464709993,
+      "grad_norm": 5.303604758140139,
+      "learning_rate": 4.961163319653959e-06,
+      "loss": 0.6175,
+      "step": 118
+    },
+    {
+      "epoch": 0.6708595387840671,
+      "grad_norm": 3.8308857949946398,
+      "learning_rate": 4.959598557219361e-06,
+      "loss": 0.6178,
+      "step": 120
+    },
+    {
+      "epoch": 0.6820405310971349,
+      "grad_norm": 5.611339241664303,
+      "learning_rate": 4.95800314809932e-06,
+      "loss": 0.617,
+      "step": 122
+    },
+    {
+      "epoch": 0.6932215234102026,
+      "grad_norm": 5.234511261826922,
+      "learning_rate": 4.956377112172691e-06,
+      "loss": 0.6557,
+      "step": 124
+    },
+    {
+      "epoch": 0.7044025157232704,
+      "grad_norm": 4.381066733905507,
+      "learning_rate": 4.954720469699939e-06,
+      "loss": 0.6343,
+      "step": 126
+    },
+    {
+      "epoch": 0.7155835080363382,
+      "grad_norm": 5.113989443684452,
+      "learning_rate": 4.953033241322887e-06,
+      "loss": 0.6135,
+      "step": 128
+    },
+    {
+      "epoch": 0.726764500349406,
+      "grad_norm": 5.138987950069777,
+      "learning_rate": 4.951315448064462e-06,
+      "loss": 0.6403,
+      "step": 130
+    },
+    {
+      "epoch": 0.7379454926624738,
+      "grad_norm": 4.43583718290579,
+      "learning_rate": 4.949567111328428e-06,
+      "loss": 0.6226,
+      "step": 132
+    },
+    {
+      "epoch": 0.7491264849755416,
+      "grad_norm": 4.391597448273059,
+      "learning_rate": 4.947788252899124e-06,
+      "loss": 0.6333,
+      "step": 134
+    },
+    {
+      "epoch": 0.7603074772886094,
+      "grad_norm": 4.193385817962468,
+      "learning_rate": 4.945978894941189e-06,
+      "loss": 0.6884,
+      "step": 136
+    },
+    {
+      "epoch": 0.7714884696016772,
+      "grad_norm": 5.03154779607414,
+      "learning_rate": 4.944139059999286e-06,
+      "loss": 0.5783,
+      "step": 138
+    },
+    {
+      "epoch": 0.782669461914745,
+      "grad_norm": 6.345004441163444,
+      "learning_rate": 4.942268770997825e-06,
+      "loss": 0.5314,
+      "step": 140
+    },
+    {
+      "epoch": 0.7938504542278128,
+      "grad_norm": 4.800013540838224,
+      "learning_rate": 4.940368051240675e-06,
+      "loss": 0.5876,
+      "step": 142
+    },
+    {
+      "epoch": 0.8050314465408805,
+      "grad_norm": 5.229387760297341,
+      "learning_rate": 4.938436924410869e-06,
+      "loss": 0.6266,
+      "step": 144
+    },
+    {
+      "epoch": 0.8162124388539483,
+      "grad_norm": 5.663117027843187,
+      "learning_rate": 4.936475414570317e-06,
+      "loss": 0.5407,
+      "step": 146
+    },
+    {
+      "epoch": 0.827393431167016,
+      "grad_norm": 4.355698674662869,
+      "learning_rate": 4.9344835461595016e-06,
+      "loss": 0.5757,
+      "step": 148
+    },
+    {
+      "epoch": 0.8385744234800838,
+      "grad_norm": 3.73012661577406,
+      "learning_rate": 4.932461343997174e-06,
+      "loss": 0.5671,
+      "step": 150
+    },
+    {
+      "epoch": 0.8497554157931516,
+      "grad_norm": 5.17610307953933,
+      "learning_rate": 4.930408833280044e-06,
+      "loss": 0.5552,
+      "step": 152
+    },
+    {
+      "epoch": 0.8609364081062194,
+      "grad_norm": 4.8108290286110575,
+      "learning_rate": 4.928326039582468e-06,
+      "loss": 0.5455,
+      "step": 154
+    },
+    {
+      "epoch": 0.8721174004192872,
+      "grad_norm": 4.143977047297293,
+      "learning_rate": 4.926212988856131e-06,
+      "loss": 0.5865,
+      "step": 156
+    },
+    {
+      "epoch": 0.883298392732355,
+      "grad_norm": 4.809016102192773,
+      "learning_rate": 4.9240697074297205e-06,
+      "loss": 0.5904,
+      "step": 158
+    },
+    {
+      "epoch": 0.8944793850454228,
+      "grad_norm": 4.329310274878485,
+      "learning_rate": 4.921896222008598e-06,
+      "loss": 0.5213,
+      "step": 160
+    },
+    {
+      "epoch": 0.9056603773584906,
+      "grad_norm": 6.082276125346202,
+      "learning_rate": 4.919692559674469e-06,
+      "loss": 0.5321,
+      "step": 162
+    },
+    {
+      "epoch": 0.9168413696715584,
+      "grad_norm": 3.595682377289556,
+      "learning_rate": 4.917458747885045e-06,
+      "loss": 0.5589,
+      "step": 164
+    },
+    {
+      "epoch": 0.9280223619846262,
+      "grad_norm": 4.759398027424621,
+      "learning_rate": 4.9151948144737e-06,
+      "loss": 0.5252,
+      "step": 166
+    },
+    {
+      "epoch": 0.939203354297694,
+      "grad_norm": 4.925856740501272,
+      "learning_rate": 4.912900787649124e-06,
+      "loss": 0.5688,
+      "step": 168
+    },
+    {
+      "epoch": 0.9503843466107617,
+      "grad_norm": 4.9751554778931695,
+      "learning_rate": 4.910576695994976e-06,
+      "loss": 0.49,
+      "step": 170
+    },
+    {
+      "epoch": 0.9615653389238294,
+      "grad_norm": 4.404002437196143,
+      "learning_rate": 4.908222568469516e-06,
+      "loss": 0.5031,
+      "step": 172
+    },
+    {
+      "epoch": 0.9727463312368972,
+      "grad_norm": 4.438458089119356,
+      "learning_rate": 4.905838434405259e-06,
+      "loss": 0.5015,
+      "step": 174
+    },
+    {
+      "epoch": 0.983927323549965,
+      "grad_norm": 3.7675300141289205,
+      "learning_rate": 4.903424323508601e-06,
+      "loss": 0.5133,
+      "step": 176
+    },
+    {
+      "epoch": 0.9951083158630328,
+      "grad_norm": 5.557474516168906,
+      "learning_rate": 4.900980265859449e-06,
+      "loss": 0.4913,
+      "step": 178
+    },
+    {
+      "epoch": 1.0062893081761006,
+      "grad_norm": 4.4806858821540585,
+      "learning_rate": 4.898506291910847e-06,
+      "loss": 0.4446,
+      "step": 180
+    },
+    {
+      "epoch": 1.0174703004891683,
+      "grad_norm": 4.605929975666356,
+      "learning_rate": 4.896002432488599e-06,
+      "loss": 0.3632,
+      "step": 182
+    },
+    {
+      "epoch": 1.0286512928022362,
+      "grad_norm": 4.9794341930411665,
+      "learning_rate": 4.893468718790883e-06,
+      "loss": 0.3868,
+      "step": 184
+    },
+    {
+      "epoch": 1.039832285115304,
+      "grad_norm": 3.5317296745452733,
+      "learning_rate": 4.890905182387862e-06,
+      "loss": 0.4334,
+      "step": 186
+    },
+    {
+      "epoch": 1.0510132774283718,
+      "grad_norm": 4.568181420141649,
+      "learning_rate": 4.88831185522129e-06,
+      "loss": 0.456,
+      "step": 188
+    },
+    {
+      "epoch": 1.0621942697414395,
+      "grad_norm": 3.570260813698039,
+      "learning_rate": 4.885688769604115e-06,
+      "loss": 0.3846,
+      "step": 190
+    },
+    {
+      "epoch": 1.0733752620545074,
+      "grad_norm": 3.639759353451614,
+      "learning_rate": 4.883035958220077e-06,
+      "loss": 0.4363,
+      "step": 192
+    },
+    {
+      "epoch": 1.084556254367575,
+      "grad_norm": 4.074741691986429,
+      "learning_rate": 4.8803534541233016e-06,
+      "loss": 0.3782,
+      "step": 194
+    },
+    {
+      "epoch": 1.095737246680643,
+      "grad_norm": 4.875221867832197,
+      "learning_rate": 4.8776412907378845e-06,
+      "loss": 0.3815,
+      "step": 196
+    },
+    {
+      "epoch": 1.1069182389937107,
+      "grad_norm": 3.575182053435755,
+      "learning_rate": 4.874899501857477e-06,
+      "loss": 0.4023,
+      "step": 198
+    },
+    {
+      "epoch": 1.1180992313067786,
+      "grad_norm": 3.984785984285916,
+      "learning_rate": 4.8721281216448675e-06,
+      "loss": 0.305,
+      "step": 200
+    },
+    {
+      "epoch": 1.1292802236198463,
+      "grad_norm": 3.997235184408756,
+      "learning_rate": 4.869327184631552e-06,
+      "loss": 0.3896,
+      "step": 202
+    },
+    {
+      "epoch": 1.140461215932914,
+      "grad_norm": 3.403723018382878,
+      "learning_rate": 4.866496725717304e-06,
+      "loss": 0.3332,
+      "step": 204
+    },
+    {
+      "epoch": 1.1516422082459818,
+      "grad_norm": 3.5740869992425917,
+      "learning_rate": 4.8636367801697415e-06,
+      "loss": 0.3299,
+      "step": 206
+    },
+    {
+      "epoch": 1.1628232005590495,
+      "grad_norm": 3.8789874672120033,
+      "learning_rate": 4.860747383623889e-06,
+      "loss": 0.4145,
+      "step": 208
+    },
+    {
+      "epoch": 1.1740041928721174,
+      "grad_norm": 3.8038820435820084,
+      "learning_rate": 4.857828572081731e-06,
+      "loss": 0.3171,
+      "step": 210
+    },
+    {
+      "epoch": 1.1851851851851851,
+      "grad_norm": 3.260333619392394,
+      "learning_rate": 4.854880381911762e-06,
+      "loss": 0.3474,
+      "step": 212
+    },
+    {
+      "epoch": 1.196366177498253,
+      "grad_norm": 2.8989963280714925,
+      "learning_rate": 4.851902849848536e-06,
+      "loss": 0.3931,
+      "step": 214
+    },
+    {
+      "epoch": 1.2075471698113207,
+      "grad_norm": 3.6383247911373773,
+      "learning_rate": 4.848896012992208e-06,
+      "loss": 0.3822,
+      "step": 216
+    },
+    {
+      "epoch": 1.2187281621243886,
+      "grad_norm": 3.0864181531286734,
+      "learning_rate": 4.845859908808074e-06,
+      "loss": 0.378,
+      "step": 218
+    },
+    {
+      "epoch": 1.2299091544374563,
+      "grad_norm": 2.494513481207721,
+      "learning_rate": 4.842794575126099e-06,
+      "loss": 0.3655,
+      "step": 220
+    },
+    {
+      "epoch": 1.2410901467505242,
+      "grad_norm": 2.6074910342756334,
+      "learning_rate": 4.839700050140448e-06,
+      "loss": 0.3973,
+      "step": 222
+    },
+    {
+      "epoch": 1.2522711390635919,
+      "grad_norm": 2.2421870374103285,
+      "learning_rate": 4.836576372409015e-06,
+      "loss": 0.3784,
+      "step": 224
+    },
+    {
+      "epoch": 1.2634521313766598,
+      "grad_norm": 2.451559449193117,
+      "learning_rate": 4.833423580852933e-06,
+      "loss": 0.3805,
+      "step": 226
+    },
+    {
+      "epoch": 1.2746331236897275,
+      "grad_norm": 2.5374184019501285,
+      "learning_rate": 4.830241714756099e-06,
+      "loss": 0.293,
+      "step": 228
+    },
+    {
+      "epoch": 1.2858141160027952,
+      "grad_norm": 2.525807489259318,
+      "learning_rate": 4.827030813764677e-06,
+      "loss": 0.2665,
+      "step": 230
+    },
+    {
+      "epoch": 1.296995108315863,
+      "grad_norm": 2.3755504317471523,
+      "learning_rate": 4.8237909178866075e-06,
+      "loss": 0.4108,
+      "step": 232
+    },
+    {
+      "epoch": 1.3081761006289307,
+      "grad_norm": 2.7662660096000793,
+      "learning_rate": 4.8205220674911075e-06,
+      "loss": 0.3928,
+      "step": 234
+    },
+    {
+      "epoch": 1.3193570929419987,
+      "grad_norm": 2.245517906271987,
+      "learning_rate": 4.81722430330817e-06,
+      "loss": 0.355,
+      "step": 236
+    },
+    {
+      "epoch": 1.3305380852550663,
+      "grad_norm": 2.684087860818518,
+      "learning_rate": 4.813897666428054e-06,
+      "loss": 0.3624,
+      "step": 238
+    },
+    {
+      "epoch": 1.3417190775681342,
+      "grad_norm": 2.5507370157459865,
+      "learning_rate": 4.810542198300772e-06,
+      "loss": 0.3494,
+      "step": 240
+    },
+    {
+      "epoch": 1.352900069881202,
+      "grad_norm": 2.157612559104276,
+      "learning_rate": 4.807157940735577e-06,
+      "loss": 0.3064,
+      "step": 242
+    },
+    {
+      "epoch": 1.3640810621942698,
+      "grad_norm": 1.9389355017962189,
+      "learning_rate": 4.803744935900439e-06,
+      "loss": 0.3331,
+      "step": 244
+    },
+    {
+      "epoch": 1.3752620545073375,
+      "grad_norm": 2.3147558047608867,
+      "learning_rate": 4.8003032263215185e-06,
+      "loss": 0.3538,
+      "step": 246
+    },
+    {
+      "epoch": 1.3864430468204052,
+      "grad_norm": 2.414181223767401,
+      "learning_rate": 4.79683285488264e-06,
+      "loss": 0.3237,
+      "step": 248
+    },
+    {
+      "epoch": 1.397624039133473,
+      "grad_norm": 2.0498128676624368,
+      "learning_rate": 4.793333864824756e-06,
+      "loss": 0.3742,
+      "step": 250
+    },
+    {
+      "epoch": 1.408805031446541,
+      "grad_norm": 2.2294049255917416,
+      "learning_rate": 4.789806299745405e-06,
+      "loss": 0.2948,
+      "step": 252
+    },
+    {
+      "epoch": 1.4199860237596087,
+      "grad_norm": 2.2210196470155923,
+      "learning_rate": 4.786250203598174e-06,
+      "loss": 0.28,
+      "step": 254
+    },
+    {
+      "epoch": 1.4311670160726764,
+      "grad_norm": 2.6896787603814816,
+      "learning_rate": 4.782665620692147e-06,
+      "loss": 0.3513,
+      "step": 256
+    },
+    {
+      "epoch": 1.4423480083857443,
+      "grad_norm": 2.1151921249556644,
+      "learning_rate": 4.779052595691355e-06,
+      "loss": 0.3598,
+      "step": 258
+    },
+    {
+      "epoch": 1.453529000698812,
+      "grad_norm": 2.6404538176276047,
+      "learning_rate": 4.775411173614218e-06,
+      "loss": 0.3075,
+      "step": 260
+    },
+    {
+      "epoch": 1.4647099930118799,
+      "grad_norm": 1.9888888421343762,
+      "learning_rate": 4.771741399832984e-06,
+      "loss": 0.356,
+      "step": 262
+    },
+    {
+      "epoch": 1.4758909853249476,
+      "grad_norm": 2.284642426340359,
+      "learning_rate": 4.768043320073165e-06,
+      "loss": 0.2765,
+      "step": 264
+    },
+    {
+      "epoch": 1.4870719776380152,
+      "grad_norm": 2.135563450656965,
+      "learning_rate": 4.764316980412966e-06,
+      "loss": 0.2825,
+      "step": 266
+    },
+    {
+      "epoch": 1.4982529699510831,
+      "grad_norm": 1.8267552790003188,
+      "learning_rate": 4.7605624272827125e-06,
+      "loss": 0.3915,
+      "step": 268
+    },
+    {
+      "epoch": 1.509433962264151,
+      "grad_norm": 2.26569092336033,
+      "learning_rate": 4.75677970746427e-06,
+      "loss": 0.3859,
+      "step": 270
+    },
+    {
+      "epoch": 1.5206149545772187,
+      "grad_norm": 2.3510908940666346,
+      "learning_rate": 4.75296886809046e-06,
+      "loss": 0.312,
+      "step": 272
+    },
+    {
+      "epoch": 1.5317959468902864,
+      "grad_norm": 2.1562478846600883,
+      "learning_rate": 4.749129956644477e-06,
+      "loss": 0.4398,
+      "step": 274
+    },
+    {
+      "epoch": 1.5429769392033543,
+      "grad_norm": 2.1811966726037655,
+      "learning_rate": 4.745263020959296e-06,
+      "loss": 0.3221,
+      "step": 276
+    },
+    {
+      "epoch": 1.5541579315164222,
+      "grad_norm": 2.035643810106488,
+      "learning_rate": 4.741368109217072e-06,
+      "loss": 0.3317,
+      "step": 278
+    },
+    {
+      "epoch": 1.56533892382949,
+      "grad_norm": 2.0722038381676824,
+      "learning_rate": 4.737445269948543e-06,
+      "loss": 0.4627,
+      "step": 280
+    },
+    {
+      "epoch": 1.5765199161425576,
+      "grad_norm": 2.2584403073433212,
+      "learning_rate": 4.733494552032426e-06,
+      "loss": 0.352,
+      "step": 282
+    },
+    {
+      "epoch": 1.5877009084556253,
+      "grad_norm": 3.1127410509937783,
+      "learning_rate": 4.729516004694808e-06,
+      "loss": 0.3109,
+      "step": 284
+    },
+    {
+      "epoch": 1.5988819007686932,
+      "grad_norm": 1.6930738402579835,
+      "learning_rate": 4.725509677508528e-06,
+      "loss": 0.3723,
+      "step": 286
+    },
+    {
+      "epoch": 1.610062893081761,
+      "grad_norm": 2.6225330496610573,
+      "learning_rate": 4.721475620392567e-06,
+      "loss": 0.2853,
+      "step": 288
+    },
+    {
+      "epoch": 1.6212438853948288,
+      "grad_norm": 1.998954970455011,
+      "learning_rate": 4.71741388361142e-06,
+      "loss": 0.323,
+      "step": 290
+    },
+    {
+      "epoch": 1.6324248777078965,
+      "grad_norm": 2.3952745413220677,
+      "learning_rate": 4.713324517774471e-06,
+      "loss": 0.4057,
+      "step": 292
+    },
+    {
+      "epoch": 1.6436058700209644,
+      "grad_norm": 1.7339961999135642,
+      "learning_rate": 4.7092075738353625e-06,
+      "loss": 0.2855,
+      "step": 294
+    },
+    {
+      "epoch": 1.6547868623340323,
+      "grad_norm": 2.3672466509243075,
+      "learning_rate": 4.705063103091365e-06,
+      "loss": 0.277,
+      "step": 296
+    },
+    {
+      "epoch": 1.6659678546471,
+      "grad_norm": 1.92096238087282,
+      "learning_rate": 4.700891157182729e-06,
+      "loss": 0.2699,
+      "step": 298
+    },
+    {
+      "epoch": 1.6771488469601676,
+      "grad_norm": 1.6478187267877538,
+      "learning_rate": 4.696691788092049e-06,
+      "loss": 0.2875,
+      "step": 300
+    },
+    {
+      "epoch": 1.6883298392732355,
+      "grad_norm": 2.6637144089516545,
+      "learning_rate": 4.692465048143615e-06,
+      "loss": 0.3229,
+      "step": 302
+    },
+    {
+      "epoch": 1.6995108315863034,
+      "grad_norm": 2.0530281428374084,
+      "learning_rate": 4.688210990002755e-06,
+      "loss": 0.3546,
+      "step": 304
+    },
+    {
+      "epoch": 1.7106918238993711,
+      "grad_norm": 2.150198399781322,
+      "learning_rate": 4.683929666675185e-06,
+      "loss": 0.4021,
+      "step": 306
+    },
+    {
+      "epoch": 1.7218728162124388,
+      "grad_norm": 2.1752313572704542,
+      "learning_rate": 4.679621131506347e-06,
+      "loss": 0.3299,
+      "step": 308
+    },
+    {
+      "epoch": 1.7330538085255065,
+      "grad_norm": 1.9055889494341978,
+      "learning_rate": 4.6752854381807414e-06,
+      "loss": 0.2514,
+      "step": 310
+    },
+    {
+      "epoch": 1.7442348008385744,
+      "grad_norm": 2.469483649303522,
+      "learning_rate": 4.670922640721261e-06,
+      "loss": 0.332,
+      "step": 312
+    },
+    {
+      "epoch": 1.7554157931516423,
+      "grad_norm": 2.327049750502898,
+      "learning_rate": 4.666532793488518e-06,
+      "loss": 0.3482,
+      "step": 314
+    },
+    {
+      "epoch": 1.76659678546471,
+      "grad_norm": 2.0224582609864674,
+      "learning_rate": 4.662115951180164e-06,
+      "loss": 0.3192,
+      "step": 316
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 1.9568416201882894,
+      "learning_rate": 4.657672168830211e-06,
+      "loss": 0.2682,
+      "step": 318
+    },
+    {
+      "epoch": 1.7889587700908456,
+      "grad_norm": 1.919410926201314,
+      "learning_rate": 4.653201501808346e-06,
+      "loss": 0.3602,
+      "step": 320
+    },
+    {
+      "epoch": 1.8001397624039135,
+      "grad_norm": 2.239752835185363,
+      "learning_rate": 4.6487040058192385e-06,
+      "loss": 0.346,
+      "step": 322
+    },
+    {
+      "epoch": 1.8113207547169812,
+      "grad_norm": 2.3820790461811643,
+      "learning_rate": 4.644179736901848e-06,
+      "loss": 0.393,
+      "step": 324
+    },
+    {
+      "epoch": 1.8225017470300489,
+      "grad_norm": 2.100652056063807,
+      "learning_rate": 4.639628751428728e-06,
+      "loss": 0.3348,
+      "step": 326
+    },
+    {
+      "epoch": 1.8336827393431165,
+      "grad_norm": 1.839587786014522,
+      "learning_rate": 4.635051106105316e-06,
+      "loss": 0.297,
+      "step": 328
+    },
+    {
+      "epoch": 1.8448637316561844,
+      "grad_norm": 1.460937373317575,
+      "learning_rate": 4.630446857969238e-06,
+      "loss": 0.3291,
+      "step": 330
+    },
+    {
+      "epoch": 1.8560447239692524,
+      "grad_norm": 3.066440662132836,
+      "learning_rate": 4.625816064389589e-06,
+      "loss": 0.2752,
+      "step": 332
+    },
+    {
+      "epoch": 1.86722571628232,
+      "grad_norm": 1.9596525632755366,
+      "learning_rate": 4.62115878306622e-06,
+      "loss": 0.3444,
+      "step": 334
+    },
+    {
+      "epoch": 1.8784067085953877,
+      "grad_norm": 2.2835299782118335,
+      "learning_rate": 4.616475072029024e-06,
+      "loss": 0.3013,
+      "step": 336
+    },
+    {
+      "epoch": 1.8895877009084556,
+      "grad_norm": 2.1330589159921756,
+      "learning_rate": 4.6117649896372055e-06,
+      "loss": 0.3811,
+      "step": 338
+    },
+    {
+      "epoch": 1.9007686932215235,
+      "grad_norm": 2.28792058261577,
+      "learning_rate": 4.607028594578559e-06,
+      "loss": 0.304,
+      "step": 340
+    },
+    {
+      "epoch": 1.9119496855345912,
+      "grad_norm": 1.8457539990364031,
+      "learning_rate": 4.602265945868735e-06,
+      "loss": 0.2817,
+      "step": 342
+    },
+    {
+      "epoch": 1.923130677847659,
+      "grad_norm": 1.7860630390403116,
+      "learning_rate": 4.597477102850506e-06,
+      "loss": 0.3166,
+      "step": 344
+    },
+    {
+      "epoch": 1.9343116701607268,
+      "grad_norm": 1.988441202911347,
+      "learning_rate": 4.592662125193027e-06,
+      "loss": 0.2881,
+      "step": 346
+    },
+    {
+      "epoch": 1.9454926624737947,
+      "grad_norm": 1.7341207391896365,
+      "learning_rate": 4.587821072891089e-06,
+      "loss": 0.3126,
+      "step": 348
+    },
+    {
+      "epoch": 1.9566736547868624,
+      "grad_norm": 1.8960045369195677,
+      "learning_rate": 4.582954006264377e-06,
+      "loss": 0.32,
+      "step": 350
+    },
+    {
+      "epoch": 1.96785464709993,
+      "grad_norm": 1.8028316706058551,
+      "learning_rate": 4.578060985956714e-06,
+      "loss": 0.3308,
+      "step": 352
+    },
+    {
+      "epoch": 1.9790356394129978,
+      "grad_norm": 1.7537644172052635,
+      "learning_rate": 4.573142072935307e-06,
+      "loss": 0.325,
+      "step": 354
+    },
+    {
+      "epoch": 1.9902166317260657,
+      "grad_norm": 1.5291097261080726,
+      "learning_rate": 4.568197328489986e-06,
+      "loss": 0.3418,
+      "step": 356
+    },
+    {
+      "epoch": 2.0013976240391336,
+      "grad_norm": 2.703429613422267,
+      "learning_rate": 4.563226814232444e-06,
+      "loss": 0.316,
+      "step": 358
+    },
+    {
+      "epoch": 2.0125786163522013,
+      "grad_norm": 1.6677019482039983,
+      "learning_rate": 4.558230592095465e-06,
+      "loss": 0.2242,
+      "step": 360
+    },
+    {
+      "epoch": 2.023759608665269,
+      "grad_norm": 2.1855279147060527,
+      "learning_rate": 4.5532087243321536e-06,
+      "loss": 0.1706,
+      "step": 362
+    },
+    {
+      "epoch": 2.0349406009783366,
+      "grad_norm": 1.433260386596143,
+      "learning_rate": 4.548161273515161e-06,
+      "loss": 0.2597,
+      "step": 364
+    },
+    {
+      "epoch": 2.0461215932914047,
+      "grad_norm": 1.9528007044032762,
+      "learning_rate": 4.543088302535903e-06,
+      "loss": 0.2321,
+      "step": 366
+    },
+    {
+      "epoch": 2.0573025856044724,
+      "grad_norm": 1.508509476663671,
+      "learning_rate": 4.53798987460378e-06,
+      "loss": 0.1975,
+      "step": 368
+    },
+    {
+      "epoch": 2.06848357791754,
+      "grad_norm": 1.4870411030447606,
+      "learning_rate": 4.532866053245385e-06,
+      "loss": 0.218,
+      "step": 370
+    },
+    {
+      "epoch": 2.079664570230608,
+      "grad_norm": 1.984299603467917,
+      "learning_rate": 4.527716902303713e-06,
+      "loss": 0.1866,
+      "step": 372
+    },
+    {
+      "epoch": 2.090845562543676,
+      "grad_norm": 1.7502708144873231,
+      "learning_rate": 4.522542485937369e-06,
+      "loss": 0.2128,
+      "step": 374
+    },
+    {
+      "epoch": 2.1020265548567436,
+      "grad_norm": 1.131006072907252,
+      "learning_rate": 4.517342868619764e-06,
+      "loss": 0.2418,
+      "step": 376
+    },
+    {
+      "epoch": 2.1132075471698113,
+      "grad_norm": 2.365723778930082,
+      "learning_rate": 4.512118115138315e-06,
+      "loss": 0.2249,
+      "step": 378
+    },
+    {
+      "epoch": 2.124388539482879,
+      "grad_norm": 1.7739738087900154,
+      "learning_rate": 4.506868290593635e-06,
+      "loss": 0.225,
+      "step": 380
+    },
+    {
+      "epoch": 2.135569531795947,
+      "grad_norm": 2.3920039733015197,
+      "learning_rate": 4.501593460398726e-06,
+      "loss": 0.207,
+      "step": 382
+    },
+    {
+      "epoch": 2.146750524109015,
+      "grad_norm": 1.3961875749075527,
+      "learning_rate": 4.49629369027816e-06,
+      "loss": 0.1847,
+      "step": 384
+    },
+    {
+      "epoch": 2.1579315164220825,
+      "grad_norm": 1.740079266616333,
+      "learning_rate": 4.490969046267258e-06,
+      "loss": 0.2092,
+      "step": 386
+    },
+    {
+      "epoch": 2.16911250873515,
+      "grad_norm": 1.716849109423316,
+      "learning_rate": 4.485619594711278e-06,
+      "loss": 0.2512,
+      "step": 388
+    },
+    {
+      "epoch": 2.180293501048218,
+      "grad_norm": 2.2256205473256836,
+      "learning_rate": 4.4802454022645725e-06,
+      "loss": 0.2212,
+      "step": 390
+    },
+    {
+      "epoch": 2.191474493361286,
+      "grad_norm": 1.5080548485099736,
+      "learning_rate": 4.474846535889773e-06,
+      "loss": 0.2577,
+      "step": 392
+    },
+    {
+      "epoch": 2.2026554856743537,
+      "grad_norm": 1.849350001917602,
+      "learning_rate": 4.469423062856946e-06,
+      "loss": 0.2518,
+      "step": 394
+    },
+    {
+      "epoch": 2.2138364779874213,
+      "grad_norm": 2.0456903454646937,
+      "learning_rate": 4.463975050742757e-06,
+      "loss": 0.2666,
+      "step": 396
+    },
+    {
+      "epoch": 2.225017470300489,
+      "grad_norm": 2.1576955140860172,
+      "learning_rate": 4.4585025674296315e-06,
+      "loss": 0.1881,
+      "step": 398
+    },
+    {
+      "epoch": 2.236198462613557,
+      "grad_norm": 1.959825305986428,
+      "learning_rate": 4.453005681104906e-06,
+      "loss": 0.1912,
+      "step": 400
+    },
+    {
+      "epoch": 2.247379454926625,
+      "grad_norm": 1.8263078605633967,
+      "learning_rate": 4.44748446025998e-06,
+      "loss": 0.177,
+      "step": 402
+    },
+    {
+      "epoch": 2.2585604472396925,
+      "grad_norm": 1.3737693376807456,
+      "learning_rate": 4.44193897368946e-06,
+      "loss": 0.2083,
+      "step": 404
+    },
+    {
+      "epoch": 2.26974143955276,
+      "grad_norm": 1.9216745648550881,
+      "learning_rate": 4.436369290490307e-06,
+      "loss": 0.269,
+      "step": 406
+    },
+    {
+      "epoch": 2.280922431865828,
+      "grad_norm": 1.5225068983698562,
+      "learning_rate": 4.430775480060973e-06,
+      "loss": 0.2043,
+      "step": 408
+    },
+    {
+      "epoch": 2.292103424178896,
+      "grad_norm": 1.958524495155971,
+      "learning_rate": 4.425157612100531e-06,
+      "loss": 0.2735,
+      "step": 410
+    },
+    {
+      "epoch": 2.3032844164919637,
+      "grad_norm": 2.020109840115744,
+      "learning_rate": 4.419515756607819e-06,
+      "loss": 0.2623,
+      "step": 412
+    },
+    {
+      "epoch": 2.3144654088050314,
+      "grad_norm": 1.6832635446278787,
+      "learning_rate": 4.413849983880554e-06,
+      "loss": 0.2122,
+      "step": 414
+    },
+    {
+      "epoch": 2.325646401118099,
+      "grad_norm": 1.8238819367042174,
+      "learning_rate": 4.4081603645144685e-06,
+      "loss": 0.2141,
+      "step": 416
+    },
+    {
+      "epoch": 2.336827393431167,
+      "grad_norm": 1.636664838162331,
+      "learning_rate": 4.4024469694024194e-06,
+      "loss": 0.2159,
+      "step": 418
+    },
+    {
+      "epoch": 2.348008385744235,
+      "grad_norm": 1.563361723149053,
+      "learning_rate": 4.396709869733515e-06,
+      "loss": 0.2636,
+      "step": 420
+    },
+    {
+      "epoch": 2.3591893780573026,
+      "grad_norm": 1.7104549540666967,
+      "learning_rate": 4.39094913699222e-06,
+      "loss": 0.2059,
+      "step": 422
+    },
+    {
+      "epoch": 2.3703703703703702,
+      "grad_norm": 1.7448299629844894,
+      "learning_rate": 4.385164842957469e-06,
+      "loss": 0.2076,
+      "step": 424
+    },
+    {
+      "epoch": 2.381551362683438,
+      "grad_norm": 2.0760771369111812,
+      "learning_rate": 4.379357059701771e-06,
+      "loss": 0.2241,
+      "step": 426
+    },
+    {
+      "epoch": 2.392732354996506,
+      "grad_norm": 1.4610379659131663,
+      "learning_rate": 4.373525859590313e-06,
+      "loss": 0.2135,
+      "step": 428
+    },
+    {
+      "epoch": 2.4039133473095737,
+      "grad_norm": 1.9763200369365506,
+      "learning_rate": 4.367671315280055e-06,
+      "loss": 0.2225,
+      "step": 430
+    },
+    {
+      "epoch": 2.4150943396226414,
+      "grad_norm": 2.138415914668256,
+      "learning_rate": 4.3617934997188274e-06,
+      "loss": 0.2618,
+      "step": 432
+    },
+    {
+      "epoch": 2.426275331935709,
+      "grad_norm": 1.6842725394389781,
+      "learning_rate": 4.355892486144419e-06,
+      "loss": 0.1691,
+      "step": 434
+    },
+    {
+      "epoch": 2.4374563242487772,
+      "grad_norm": 2.056626946764254,
+      "learning_rate": 4.349968348083673e-06,
+      "loss": 0.1922,
+      "step": 436
+    },
+    {
+      "epoch": 2.448637316561845,
+      "grad_norm": 1.2423274511146358,
+      "learning_rate": 4.3440211593515556e-06,
+      "loss": 0.2061,
+      "step": 438
+    },
+    {
+      "epoch": 2.4598183088749126,
+      "grad_norm": 1.465237522133527,
+      "learning_rate": 4.338050994050253e-06,
+      "loss": 0.1996,
+      "step": 440
+    },
+    {
+      "epoch": 2.4709993011879803,
+      "grad_norm": 2.1451900105983315,
+      "learning_rate": 4.332057926568235e-06,
+      "loss": 0.2441,
+      "step": 442
+    },
+    {
+      "epoch": 2.4821802935010484,
+      "grad_norm": 1.5259606296511572,
+      "learning_rate": 4.326042031579337e-06,
+      "loss": 0.2066,
+      "step": 444
+    },
+    {
+      "epoch": 2.493361285814116,
+      "grad_norm": 2.4163109674867784,
+      "learning_rate": 4.320003384041823e-06,
+      "loss": 0.2393,
+      "step": 446
+    },
+    {
+      "epoch": 2.5045422781271838,
+      "grad_norm": 2.1518283309231907,
+      "learning_rate": 4.313942059197457e-06,
+      "loss": 0.2467,
+      "step": 448
+    },
+    {
+      "epoch": 2.5157232704402515,
+      "grad_norm": 1.6715387204280183,
+      "learning_rate": 4.3078581325705614e-06,
+      "loss": 0.2495,
+      "step": 450
+    },
+    {
+      "epoch": 2.5269042627533196,
+      "grad_norm": 1.7729216990478125,
+      "learning_rate": 4.3017516799670785e-06,
+      "loss": 0.1586,
+      "step": 452
+    },
+    {
+      "epoch": 2.5380852550663873,
+      "grad_norm": 1.7853923740535589,
+      "learning_rate": 4.295622777473625e-06,
+      "loss": 0.2216,
+      "step": 454
+    },
+    {
+      "epoch": 2.549266247379455,
+      "grad_norm": 1.7001940457803237,
+      "learning_rate": 4.289471501456543e-06,
+      "loss": 0.2288,
+      "step": 456
+    },
+    {
+      "epoch": 2.5604472396925226,
+      "grad_norm": 2.5868877625212354,
+      "learning_rate": 4.283297928560951e-06,
+      "loss": 0.2075,
+      "step": 458
+    },
+    {
+      "epoch": 2.5716282320055903,
+      "grad_norm": 2.1990912649669823,
+      "learning_rate": 4.277102135709786e-06,
+      "loss": 0.2017,
+      "step": 460
+    },
+    {
+      "epoch": 2.582809224318658,
+      "grad_norm": 2.2627396419665273,
+      "learning_rate": 4.270884200102848e-06,
+      "loss": 0.2144,
+      "step": 462
+    },
+    {
+      "epoch": 2.593990216631726,
+      "grad_norm": 2.2283930780278505,
+      "learning_rate": 4.2646441992158356e-06,
+      "loss": 0.3,
+      "step": 464
+    },
+    {
+      "epoch": 2.605171208944794,
+      "grad_norm": 2.6765537923336087,
+      "learning_rate": 4.258382210799381e-06,
+      "loss": 0.2441,
+      "step": 466
+    },
+    {
+      "epoch": 2.6163522012578615,
+      "grad_norm": 2.0124117535310706,
+      "learning_rate": 4.252098312878083e-06,
+      "loss": 0.2667,
+      "step": 468
+    },
+    {
+      "epoch": 2.6275331935709296,
+      "grad_norm": 2.0622543839995586,
+      "learning_rate": 4.245792583749533e-06,
+      "loss": 0.2209,
+      "step": 470
+    },
+    {
+      "epoch": 2.6387141858839973,
+      "grad_norm": 1.7479329049755916,
+      "learning_rate": 4.2394651019833385e-06,
+      "loss": 0.2045,
+      "step": 472
+    },
+    {
+      "epoch": 2.649895178197065,
+      "grad_norm": 2.223724201139868,
+      "learning_rate": 4.23311594642015e-06,
+      "loss": 0.2283,
+      "step": 474
+    },
+    {
+      "epoch": 2.6610761705101327,
+      "grad_norm": 1.8280919056271019,
+      "learning_rate": 4.226745196170669e-06,
+      "loss": 0.2319,
+      "step": 476
+    },
+    {
+      "epoch": 2.6722571628232004,
+      "grad_norm": 1.6911807333452673,
+      "learning_rate": 4.220352930614672e-06,
+      "loss": 0.232,
+      "step": 478
+    },
+    {
+      "epoch": 2.6834381551362685,
+      "grad_norm": 1.9242468593637576,
+      "learning_rate": 4.213939229400014e-06,
+      "loss": 0.2733,
+      "step": 480
+    },
+    {
+      "epoch": 2.694619147449336,
+      "grad_norm": 2.1223012349945254,
+      "learning_rate": 4.20750417244164e-06,
+      "loss": 0.2529,
+      "step": 482
+    },
+    {
+      "epoch": 2.705800139762404,
+      "grad_norm": 2.1921742273194313,
+      "learning_rate": 4.201047839920589e-06,
+      "loss": 0.257,
+      "step": 484
+    },
+    {
+      "epoch": 2.7169811320754715,
+      "grad_norm": 2.118251084662083,
+      "learning_rate": 4.194570312282993e-06,
+      "loss": 0.235,
+      "step": 486
+    },
+    {
+      "epoch": 2.7281621243885397,
+      "grad_norm": 1.9816644323530734,
+      "learning_rate": 4.1880716702390764e-06,
+      "loss": 0.1839,
+      "step": 488
+    },
+    {
+      "epoch": 2.7393431167016074,
+      "grad_norm": 1.8891363830208663,
+      "learning_rate": 4.181551994762151e-06,
+      "loss": 0.2301,
+      "step": 490
+    },
+    {
+      "epoch": 2.750524109014675,
+      "grad_norm": 1.7502840233703516,
+      "learning_rate": 4.1750113670876045e-06,
+      "loss": 0.1883,
+      "step": 492
+    },
+    {
+      "epoch": 2.7617051013277427,
+      "grad_norm": 1.5627429248705165,
+      "learning_rate": 4.16844986871189e-06,
+      "loss": 0.2042,
+      "step": 494
+    },
+    {
+      "epoch": 2.7728860936408104,
+      "grad_norm": 1.8631447011251083,
+      "learning_rate": 4.161867581391511e-06,
+      "loss": 0.2018,
+      "step": 496
+    },
+    {
+      "epoch": 2.7840670859538785,
+      "grad_norm": 2.0906363974353765,
+      "learning_rate": 4.155264587142002e-06,
+      "loss": 0.2319,
+      "step": 498
+    },
+    {
+      "epoch": 2.795248078266946,
+      "grad_norm": 1.7819164584799931,
+      "learning_rate": 4.148640968236903e-06,
+      "loss": 0.1703,
+      "step": 500
+    },
+    {
+      "epoch": 2.806429070580014,
+      "grad_norm": 1.7607086842324982,
+      "learning_rate": 4.141996807206745e-06,
+      "loss": 0.2264,
+      "step": 502
+    },
+    {
+      "epoch": 2.817610062893082,
+      "grad_norm": 1.5277530729360727,
+      "learning_rate": 4.135332186838008e-06,
+      "loss": 0.2134,
+      "step": 504
+    },
+    {
+      "epoch": 2.8287910552061497,
+      "grad_norm": 1.739277840645659,
+      "learning_rate": 4.128647190172099e-06,
+      "loss": 0.1952,
+      "step": 506
+    },
+    {
+      "epoch": 2.8399720475192174,
+      "grad_norm": 1.9987218712547774,
+      "learning_rate": 4.121941900504316e-06,
+      "loss": 0.2364,
+      "step": 508
+    },
+    {
+      "epoch": 2.851153039832285,
+      "grad_norm": 2.2244662318443225,
+      "learning_rate": 4.1152164013828035e-06,
+      "loss": 0.2072,
+      "step": 510
+    },
+    {
+      "epoch": 2.8623340321453528,
+      "grad_norm": 1.526547678145968,
+      "learning_rate": 4.108470776607521e-06,
+      "loss": 0.2047,
+      "step": 512
+    },
+    {
+      "epoch": 2.8735150244584204,
+      "grad_norm": 2.005093613185987,
+      "learning_rate": 4.1017051102291946e-06,
+      "loss": 0.2789,
+      "step": 514
+    },
+    {
+      "epoch": 2.8846960167714886,
+      "grad_norm": 2.2990829029486624,
+      "learning_rate": 4.094919486548266e-06,
+      "loss": 0.2414,
+      "step": 516
+    },
+    {
+      "epoch": 2.8958770090845563,
+      "grad_norm": 2.13743283403912,
+      "learning_rate": 4.088113990113846e-06,
+      "loss": 0.2029,
+      "step": 518
+    },
+    {
+      "epoch": 2.907058001397624,
+      "grad_norm": 1.9027626030017704,
+      "learning_rate": 4.081288705722666e-06,
+      "loss": 0.2229,
+      "step": 520
+    },
+    {
+      "epoch": 2.918238993710692,
+      "grad_norm": 2.0076859155071745,
+      "learning_rate": 4.074443718418009e-06,
+      "loss": 0.1995,
+      "step": 522
+    },
+    {
+      "epoch": 2.9294199860237597,
+      "grad_norm": 1.7985240007466619,
+      "learning_rate": 4.067579113488661e-06,
+      "loss": 0.1807,
+      "step": 524
+    },
+    {
+      "epoch": 2.9406009783368274,
+      "grad_norm": 2.140934337000471,
+      "learning_rate": 4.060694976467844e-06,
+      "loss": 0.2532,
+      "step": 526
+    },
+    {
+      "epoch": 2.951781970649895,
+      "grad_norm": 2.323003193893417,
+      "learning_rate": 4.0537913931321495e-06,
+      "loss": 0.2421,
+      "step": 528
+    },
+    {
+      "epoch": 2.962962962962963,
+      "grad_norm": 1.4532319163010707,
+      "learning_rate": 4.04686844950047e-06,
+      "loss": 0.2267,
+      "step": 530
+    },
+    {
+      "epoch": 2.9741439552760305,
+      "grad_norm": 2.0854922336923023,
+      "learning_rate": 4.039926231832931e-06,
+      "loss": 0.266,
+      "step": 532
+    },
+    {
+      "epoch": 2.9853249475890986,
+      "grad_norm": 2.882533995321225,
+      "learning_rate": 4.032964826629811e-06,
+      "loss": 0.2079,
+      "step": 534
+    },
+    {
+      "epoch": 2.9965059399021663,
+      "grad_norm": 2.7236955724192873,
+      "learning_rate": 4.025984320630465e-06,
+      "loss": 0.1657,
+      "step": 536
+    },
+    {
+      "epoch": 3.007686932215234,
+      "grad_norm": 1.8432900490614266,
+      "learning_rate": 4.018984800812248e-06,
+      "loss": 0.1354,
+      "step": 538
+    },
+    {
+      "epoch": 3.018867924528302,
+      "grad_norm": 2.0142515580054017,
+      "learning_rate": 4.011966354389424e-06,
+      "loss": 0.1542,
+      "step": 540
+    },
+    {
+      "epoch": 3.03004891684137,
+      "grad_norm": 2.756352182005047,
+      "learning_rate": 4.004929068812086e-06,
+      "loss": 0.1638,
+      "step": 542
+    },
+    {
+      "epoch": 3.0412299091544375,
+      "grad_norm": 2.048077691313813,
+      "learning_rate": 3.997873031765061e-06,
+      "loss": 0.156,
+      "step": 544
+    },
+    {
+      "epoch": 3.052410901467505,
+      "grad_norm": 1.7442233155652336,
+      "learning_rate": 3.990798331166822e-06,
+      "loss": 0.1095,
+      "step": 546
+    },
+    {
+      "epoch": 3.063591893780573,
+      "grad_norm": 1.826861973142375,
+      "learning_rate": 3.983705055168391e-06,
+      "loss": 0.1195,
+      "step": 548
+    },
+    {
+      "epoch": 3.074772886093641,
+      "grad_norm": 1.943175517862748,
+      "learning_rate": 3.976593292152238e-06,
+      "loss": 0.1638,
+      "step": 550
+    },
+    {
+      "epoch": 3.0859538784067087,
+      "grad_norm": 1.5477727978546996,
+      "learning_rate": 3.969463130731183e-06,
+      "loss": 0.1291,
+      "step": 552
+    },
+    {
+      "epoch": 3.0971348707197763,
+      "grad_norm": 2.3918080397656034,
+      "learning_rate": 3.9623146597472915e-06,
+      "loss": 0.1333,
+      "step": 554
+    },
+    {
+      "epoch": 3.108315863032844,
+      "grad_norm": 2.0592865934704,
+      "learning_rate": 3.955147968270764e-06,
+      "loss": 0.1692,
+      "step": 556
+    },
+    {
+      "epoch": 3.119496855345912,
+      "grad_norm": 1.280306245998938,
+      "learning_rate": 3.947963145598833e-06,
+      "loss": 0.1695,
+      "step": 558
+    },
+    {
+      "epoch": 3.13067784765898,
+      "grad_norm": 1.5568837418874426,
+      "learning_rate": 3.940760281254645e-06,
+      "loss": 0.1614,
+      "step": 560
+    },
+    {
+      "epoch": 3.1418588399720475,
+      "grad_norm": 1.6248982612645957,
+      "learning_rate": 3.933539464986143e-06,
+      "loss": 0.1184,
+      "step": 562
+    },
+    {
+      "epoch": 3.153039832285115,
+      "grad_norm": 1.657284019650329,
+      "learning_rate": 3.926300786764957e-06,
+      "loss": 0.1523,
+      "step": 564
+    },
+    {
+      "epoch": 3.164220824598183,
+      "grad_norm": 1.9315037734198213,
+      "learning_rate": 3.919044336785274e-06,
+      "loss": 0.1411,
+      "step": 566
+    },
+    {
+      "epoch": 3.175401816911251,
+      "grad_norm": 1.7456382044347782,
+      "learning_rate": 3.911770205462717e-06,
+      "loss": 0.1764,
+      "step": 568
+    },
+    {
+      "epoch": 3.1865828092243187,
+      "grad_norm": 1.4045398532057205,
+      "learning_rate": 3.904478483433223e-06,
+      "loss": 0.1241,
+      "step": 570
+    },
+    {
+      "epoch": 3.1977638015373864,
+      "grad_norm": 2.0886459168414895,
+      "learning_rate": 3.897169261551907e-06,
+      "loss": 0.1475,
+      "step": 572
+    },
+    {
+      "epoch": 3.208944793850454,
+      "grad_norm": 1.9098750157027404,
+      "learning_rate": 3.889842630891934e-06,
+      "loss": 0.138,
+      "step": 574
+    },
+    {
+      "epoch": 3.220125786163522,
+      "grad_norm": 2.184899827108709,
+      "learning_rate": 3.8824986827433804e-06,
+      "loss": 0.1315,
+      "step": 576
+    },
+    {
+      "epoch": 3.23130677847659,
+      "grad_norm": 1.528868394326383,
+      "learning_rate": 3.875137508612104e-06,
+      "loss": 0.1447,
+      "step": 578
+    },
+    {
+      "epoch": 3.2424877707896576,
+      "grad_norm": 1.6893708687857107,
+      "learning_rate": 3.867759200218594e-06,
+      "loss": 0.1746,
+      "step": 580
+    },
+    {
+      "epoch": 3.2536687631027252,
+      "grad_norm": 1.2610411246909474,
+      "learning_rate": 3.860363849496836e-06,
+      "loss": 0.1301,
+      "step": 582
+    },
+    {
+      "epoch": 3.264849755415793,
+      "grad_norm": 1.397542140556738,
+      "learning_rate": 3.852951548593161e-06,
+      "loss": 0.1373,
+      "step": 584
+    },
+    {
+      "epoch": 3.276030747728861,
+      "grad_norm": 1.9903353672741917,
+      "learning_rate": 3.845522389865106e-06,
+      "loss": 0.1609,
+      "step": 586
+    },
+    {
+      "epoch": 3.2872117400419287,
+      "grad_norm": 1.8370941337314268,
+      "learning_rate": 3.838076465880248e-06,
+      "loss": 0.148,
+      "step": 588
+    },
+    {
+      "epoch": 3.2983927323549964,
+      "grad_norm": 2.058865100613852,
+      "learning_rate": 3.830613869415069e-06,
+      "loss": 0.1483,
+      "step": 590
+    },
+    {
+      "epoch": 3.309573724668064,
+      "grad_norm": 1.5232253694216566,
+      "learning_rate": 3.823134693453782e-06,
+      "loss": 0.1621,
+      "step": 592
+    },
+    {
+      "epoch": 3.3207547169811322,
+      "grad_norm": 1.4993049111722665,
+      "learning_rate": 3.8156390311871885e-06,
+      "loss": 0.1433,
+      "step": 594
+    },
+    {
+      "epoch": 3.3319357092942,
+      "grad_norm": 1.555934394379587,
+      "learning_rate": 3.808126976011505e-06,
+      "loss": 0.1426,
+      "step": 596
+    },
+    {
+      "epoch": 3.3431167016072676,
+      "grad_norm": 1.3356473446523094,
+      "learning_rate": 3.8005986215272056e-06,
+      "loss": 0.1706,
+      "step": 598
+    },
+    {
+      "epoch": 3.3542976939203353,
+      "grad_norm": 1.9137688829035275,
+      "learning_rate": 3.7930540615378565e-06,
+      "loss": 0.1268,
+      "step": 600
+    },
+    {
+      "epoch": 3.3654786862334034,
+      "grad_norm": 1.5344748040953766,
+      "learning_rate": 3.785493390048942e-06,
+      "loss": 0.1458,
+      "step": 602
+    },
+    {
+      "epoch": 3.376659678546471,
+      "grad_norm": 1.602087497610558,
+      "learning_rate": 3.777916701266699e-06,
+      "loss": 0.1697,
+      "step": 604
+    },
+    {
+      "epoch": 3.3878406708595388,
+      "grad_norm": 1.4842568873334896,
+      "learning_rate": 3.7703240895969373e-06,
+      "loss": 0.1519,
+      "step": 606
+    },
+    {
+      "epoch": 3.3990216631726065,
+      "grad_norm": 1.53860971256147,
+      "learning_rate": 3.7627156496438686e-06,
+      "loss": 0.1691,
+      "step": 608
+    },
+    {
+      "epoch": 3.4102026554856746,
+      "grad_norm": 1.4193083610134813,
+      "learning_rate": 3.755091476208925e-06,
+      "loss": 0.1211,
+      "step": 610
+    },
+    {
+      "epoch": 3.4213836477987423,
+      "grad_norm": 1.8053625548432577,
+      "learning_rate": 3.7474516642895804e-06,
+      "loss": 0.131,
+      "step": 612
+    },
+    {
+      "epoch": 3.43256464011181,
+      "grad_norm": 1.9235537907938398,
+      "learning_rate": 3.7397963090781606e-06,
+      "loss": 0.163,
+      "step": 614
+    },
+    {
+      "epoch": 3.4437456324248776,
+      "grad_norm": 1.6022979215271898,
+      "learning_rate": 3.732125505960665e-06,
+      "loss": 0.1479,
+      "step": 616
+    },
+    {
+      "epoch": 3.4549266247379453,
+      "grad_norm": 1.663918706474492,
+      "learning_rate": 3.7244393505155713e-06,
+      "loss": 0.1376,
+      "step": 618
+    },
+    {
+      "epoch": 3.4661076170510134,
+      "grad_norm": 1.7974067820999995,
+      "learning_rate": 3.716737938512651e-06,
+      "loss": 0.1281,
+      "step": 620
+    },
+    {
+      "epoch": 3.477288609364081,
+      "grad_norm": 2.10108609081228,
+      "learning_rate": 3.709021365911772e-06,
+      "loss": 0.1388,
+      "step": 622
+    },
+    {
+      "epoch": 3.488469601677149,
+      "grad_norm": 1.367826215107555,
+      "learning_rate": 3.701289728861701e-06,
+      "loss": 0.1191,
+      "step": 624
+    },
+    {
+      "epoch": 3.4996505939902165,
+      "grad_norm": 1.7959553374302317,
+      "learning_rate": 3.693543123698913e-06,
+      "loss": 0.1758,
+      "step": 626
+    },
+    {
+      "epoch": 3.5108315863032846,
+      "grad_norm": 1.7389366148854988,
+      "learning_rate": 3.6857816469463806e-06,
+      "loss": 0.1405,
+      "step": 628
+    },
+    {
+      "epoch": 3.5220125786163523,
+      "grad_norm": 2.871162474790627,
+      "learning_rate": 3.6780053953123836e-06,
+      "loss": 0.1549,
+      "step": 630
+    },
+    {
+      "epoch": 3.53319357092942,
+      "grad_norm": 1.478751565339363,
+      "learning_rate": 3.6702144656892907e-06,
+      "loss": 0.1759,
+      "step": 632
+    },
+    {
+      "epoch": 3.5443745632424877,
+      "grad_norm": 1.4974413518112613,
+      "learning_rate": 3.662408955152364e-06,
+      "loss": 0.1078,
+      "step": 634
+    },
+    {
+      "epoch": 3.5555555555555554,
+      "grad_norm": 1.7006067350332152,
+      "learning_rate": 3.6545889609585405e-06,
+      "loss": 0.1427,
+      "step": 636
+    },
+    {
+      "epoch": 3.5667365478686235,
+      "grad_norm": 1.8754398825641954,
+      "learning_rate": 3.6467545805452266e-06,
+      "loss": 0.1893,
+      "step": 638
+    },
+    {
+      "epoch": 3.577917540181691,
+      "grad_norm": 1.7762501705151392,
+      "learning_rate": 3.6389059115290813e-06,
+      "loss": 0.1109,
+      "step": 640
+    },
+    {
+      "epoch": 3.589098532494759,
+      "grad_norm": 2.0251975300449327,
+      "learning_rate": 3.631043051704799e-06,
+      "loss": 0.121,
+      "step": 642
+    },
+    {
+      "epoch": 3.6002795248078265,
+      "grad_norm": 1.3531681902278672,
+      "learning_rate": 3.6231660990438922e-06,
+      "loss": 0.1348,
+      "step": 644
+    },
+    {
+      "epoch": 3.6114605171208947,
+      "grad_norm": 1.9724391202631109,
+      "learning_rate": 3.615275151693471e-06,
+      "loss": 0.1449,
+      "step": 646
+    },
+    {
+      "epoch": 3.6226415094339623,
+      "grad_norm": 1.785158595271644,
+      "learning_rate": 3.6073703079750204e-06,
+      "loss": 0.1485,
+      "step": 648
+    },
+    {
+      "epoch": 3.63382250174703,
+      "grad_norm": 1.829166278099355,
+      "learning_rate": 3.5994516663831734e-06,
+      "loss": 0.1192,
+      "step": 650
+    },
+    {
+      "epoch": 3.6450034940600977,
+      "grad_norm": 1.9222881871208803,
+      "learning_rate": 3.591519325584487e-06,
+      "loss": 0.1635,
+      "step": 652
+    },
+    {
+      "epoch": 3.6561844863731654,
+      "grad_norm": 2.052453811112636,
+      "learning_rate": 3.583573384416209e-06,
+      "loss": 0.1561,
+      "step": 654
+    },
+    {
+      "epoch": 3.6673654786862335,
+      "grad_norm": 1.9190051036571132,
+      "learning_rate": 3.575613941885047e-06,
+      "loss": 0.1051,
+      "step": 656
+    },
+    {
+      "epoch": 3.678546470999301,
+      "grad_norm": 1.4736638642637576,
+      "learning_rate": 3.5676410971659404e-06,
+      "loss": 0.123,
+      "step": 658
+    },
+    {
+      "epoch": 3.689727463312369,
+      "grad_norm": 1.7325761695268906,
+      "learning_rate": 3.5596549496008165e-06,
+      "loss": 0.1446,
+      "step": 660
+    },
+    {
+      "epoch": 3.700908455625437,
+      "grad_norm": 2.0344810615726288,
+      "learning_rate": 3.551655598697358e-06,
+      "loss": 0.1629,
+      "step": 662
+    },
+    {
+      "epoch": 3.7120894479385047,
+      "grad_norm": 1.936581123166174,
+      "learning_rate": 3.54364314412776e-06,
+      "loss": 0.1569,
+      "step": 664
+    },
+    {
+      "epoch": 3.7232704402515724,
+      "grad_norm": 1.3525874354992642,
+      "learning_rate": 3.535617685727494e-06,
+      "loss": 0.1082,
+      "step": 666
+    },
+    {
+      "epoch": 3.73445143256464,
+      "grad_norm": 1.6514309403224916,
+      "learning_rate": 3.527579323494055e-06,
+      "loss": 0.1431,
+      "step": 668
+    },
+    {
+      "epoch": 3.7456324248777078,
+      "grad_norm": 1.8602451468342234,
+      "learning_rate": 3.5195281575857228e-06,
+      "loss": 0.1639,
+      "step": 670
+    },
+    {
+      "epoch": 3.7568134171907754,
+      "grad_norm": 1.4731268992440232,
+      "learning_rate": 3.511464288320311e-06,
+      "loss": 0.1271,
+      "step": 672
+    },
+    {
+      "epoch": 3.7679944095038436,
+      "grad_norm": 1.37724516129253,
+      "learning_rate": 3.503387816173916e-06,
+      "loss": 0.1597,
+      "step": 674
+    },
+    {
+      "epoch": 3.7791754018169113,
+      "grad_norm": 1.7200144334067748,
+      "learning_rate": 3.495298841779669e-06,
+      "loss": 0.117,
+      "step": 676
+    },
+    {
+      "epoch": 3.790356394129979,
+      "grad_norm": 1.92538314164391,
+      "learning_rate": 3.4871974659264786e-06,
+      "loss": 0.1584,
+      "step": 678
+    },
+    {
+      "epoch": 3.801537386443047,
+      "grad_norm": 1.4718208788605616,
+      "learning_rate": 3.4790837895577752e-06,
+      "loss": 0.1333,
+      "step": 680
+    },
+    {
+      "epoch": 3.8127183787561147,
+      "grad_norm": 1.5582481918696203,
+      "learning_rate": 3.470957913770255e-06,
+      "loss": 0.1464,
+      "step": 682
+    },
+    {
+      "epoch": 3.8238993710691824,
+      "grad_norm": 1.4618275028428347,
+      "learning_rate": 3.462819939812618e-06,
+      "loss": 0.0995,
+      "step": 684
+    },
+    {
+      "epoch": 3.83508036338225,
+      "grad_norm": 1.3366351935592664,
+      "learning_rate": 3.4546699690843123e-06,
+      "loss": 0.1204,
+      "step": 686
+    },
+    {
+      "epoch": 3.846261355695318,
+      "grad_norm": 1.3780079667316787,
+      "learning_rate": 3.446508103134259e-06,
+      "loss": 0.1701,
+      "step": 688
+    },
+    {
+      "epoch": 3.8574423480083855,
+      "grad_norm": 1.7451718870626607,
+      "learning_rate": 3.4383344436595992e-06,
+      "loss": 0.1158,
+      "step": 690
+    },
+    {
+      "epoch": 3.8686233403214536,
+      "grad_norm": 2.019474198008684,
+      "learning_rate": 3.430149092504422e-06,
+      "loss": 0.1304,
+      "step": 692
+    },
+    {
+      "epoch": 3.8798043326345213,
+      "grad_norm": 1.6820935429062616,
+      "learning_rate": 3.4219521516584912e-06,
+      "loss": 0.1334,
+      "step": 694
+    },
+    {
+      "epoch": 3.890985324947589,
+      "grad_norm": 2.2578057319721236,
+      "learning_rate": 3.4137437232559834e-06,
+      "loss": 0.1557,
+      "step": 696
+    },
+    {
+      "epoch": 3.902166317260657,
+      "grad_norm": 1.3610116271561221,
+      "learning_rate": 3.4055239095742067e-06,
+      "loss": 0.1644,
+      "step": 698
+    },
+    {
+      "epoch": 3.913347309573725,
+      "grad_norm": 1.3397050224861815,
+      "learning_rate": 3.3972928130323322e-06,
+      "loss": 0.1471,
+      "step": 700
+    },
+    {
+      "epoch": 3.9245283018867925,
+      "grad_norm": 1.5234658664307734,
+      "learning_rate": 3.3890505361901153e-06,
+      "loss": 0.1195,
+      "step": 702
+    },
+    {
+      "epoch": 3.93570929419986,
+      "grad_norm": 1.763362220735128,
+      "learning_rate": 3.380797181746619e-06,
+      "loss": 0.1363,
+      "step": 704
+    },
+    {
+      "epoch": 3.946890286512928,
+      "grad_norm": 2.038986301246902,
+      "learning_rate": 3.3725328525389324e-06,
+      "loss": 0.1203,
+      "step": 706
+    },
+    {
+      "epoch": 3.958071278825996,
+      "grad_norm": 1.9046513315579439,
+      "learning_rate": 3.364257651540891e-06,
+      "loss": 0.1578,
+      "step": 708
+    },
+    {
+      "epoch": 3.9692522711390636,
+      "grad_norm": 1.423399143627221,
+      "learning_rate": 3.355971681861794e-06,
+      "loss": 0.1211,
+      "step": 710
+    },
+    {
+      "epoch": 3.9804332634521313,
+      "grad_norm": 1.5586817639667492,
+      "learning_rate": 3.3476750467451176e-06,
+      "loss": 0.153,
+      "step": 712
+    },
+    {
+      "epoch": 3.991614255765199,
+      "grad_norm": 1.4814888460752178,
+      "learning_rate": 3.33936784956723e-06,
+      "loss": 0.1288,
+      "step": 714
+    },
+    {
+      "epoch": 4.002795248078267,
+      "grad_norm": 1.6561127976965244,
+      "learning_rate": 3.331050193836104e-06,
+      "loss": 0.1196,
+      "step": 716
+    },
+    {
+      "epoch": 4.013976240391335,
+      "grad_norm": 1.8246755797846792,
+      "learning_rate": 3.322722183190025e-06,
+      "loss": 0.0983,
+      "step": 718
+    },
+    {
+      "epoch": 4.0251572327044025,
+      "grad_norm": 1.2508646883720782,
+      "learning_rate": 3.3143839213963026e-06,
+      "loss": 0.1132,
+      "step": 720
+    },
+    {
+      "epoch": 4.03633822501747,
+      "grad_norm": 1.3174073933660169,
+      "learning_rate": 3.306035512349974e-06,
+      "loss": 0.0886,
+      "step": 722
+    },
+    {
+      "epoch": 4.047519217330538,
+      "grad_norm": 1.4006843207756257,
+      "learning_rate": 3.297677060072513e-06,
+      "loss": 0.0907,
+      "step": 724
+    },
+    {
+      "epoch": 4.058700209643606,
+      "grad_norm": 2.147633002379955,
+      "learning_rate": 3.2893086687105324e-06,
+      "loss": 0.0814,
+      "step": 726
+    },
+    {
+      "epoch": 4.069881201956673,
+      "grad_norm": 1.8499679148666142,
+      "learning_rate": 3.280930442534486e-06,
+      "loss": 0.0916,
+      "step": 728
+    },
+    {
+      "epoch": 4.081062194269742,
+      "grad_norm": 1.5576608674855401,
+      "learning_rate": 3.272542485937369e-06,
+      "loss": 0.0814,
+      "step": 730
+    },
+    {
+      "epoch": 4.0922431865828095,
+      "grad_norm": 1.5258204722757824,
+      "learning_rate": 3.264144903433419e-06,
+      "loss": 0.0929,
+      "step": 732
+    },
+    {
+      "epoch": 4.103424178895877,
+      "grad_norm": 1.2377371189448831,
+      "learning_rate": 3.2557377996568135e-06,
+      "loss": 0.0933,
+      "step": 734
+    },
+    {
+      "epoch": 4.114605171208945,
+      "grad_norm": 1.6706792363129992,
+      "learning_rate": 3.247321279360363e-06,
+      "loss": 0.0957,
+      "step": 736
+    },
+    {
+      "epoch": 4.1257861635220126,
+      "grad_norm": 1.5205095000978939,
+      "learning_rate": 3.238895447414211e-06,
+      "loss": 0.1094,
+      "step": 738
+    },
+    {
+      "epoch": 4.13696715583508,
+      "grad_norm": 1.8218111131497405,
+      "learning_rate": 3.2304604088045206e-06,
+      "loss": 0.0866,
+      "step": 740
+    },
+    {
+      "epoch": 4.148148148148148,
+      "grad_norm": 1.5060146063158792,
+      "learning_rate": 3.222016268632175e-06,
+      "loss": 0.0974,
+      "step": 742
+    },
+    {
+      "epoch": 4.159329140461216,
+      "grad_norm": 2.33394735696618,
+      "learning_rate": 3.2135631321114603e-06,
+      "loss": 0.0767,
+      "step": 744
+    },
+    {
+      "epoch": 4.170510132774284,
+      "grad_norm": 1.8304481485687374,
+      "learning_rate": 3.2051011045687574e-06,
+      "loss": 0.1027,
+      "step": 746
+    },
+    {
+      "epoch": 4.181691125087352,
+      "grad_norm": 1.4496933516097028,
+      "learning_rate": 3.196630291441231e-06,
+      "loss": 0.073,
+      "step": 748
+    },
+    {
+      "epoch": 4.1928721174004195,
+      "grad_norm": 1.5989097781751378,
+      "learning_rate": 3.1881507982755126e-06,
+      "loss": 0.074,
+      "step": 750
+    },
+    {
+      "epoch": 4.204053109713487,
+      "grad_norm": 1.5479651084913313,
+      "learning_rate": 3.17966273072639e-06,
+      "loss": 0.0941,
+      "step": 752
+    },
+    {
+      "epoch": 4.215234102026555,
+      "grad_norm": 1.4844971201883568,
+      "learning_rate": 3.1711661945554857e-06,
+      "loss": 0.1171,
+      "step": 754
+    },
+    {
+      "epoch": 4.226415094339623,
+      "grad_norm": 1.538555100844062,
+      "learning_rate": 3.162661295629942e-06,
+      "loss": 0.0839,
+      "step": 756
+    },
+    {
+      "epoch": 4.23759608665269,
+      "grad_norm": 1.511356916861757,
+      "learning_rate": 3.154148139921102e-06,
+      "loss": 0.1039,
+      "step": 758
+    },
+    {
+      "epoch": 4.248777078965758,
+      "grad_norm": 1.811476489190878,
+      "learning_rate": 3.1456268335031886e-06,
+      "loss": 0.0794,
+      "step": 760
+    },
+    {
+      "epoch": 4.259958071278826,
+      "grad_norm": 1.6229333309674812,
+      "learning_rate": 3.137097482551983e-06,
+      "loss": 0.1152,
+      "step": 762
+    },
+    {
+      "epoch": 4.271139063591894,
+      "grad_norm": 1.4723017587041405,
+      "learning_rate": 3.128560193343501e-06,
+      "loss": 0.0944,
+      "step": 764
+    },
+    {
+      "epoch": 4.282320055904962,
+      "grad_norm": 1.0034690245189755,
+      "learning_rate": 3.1200150722526693e-06,
+      "loss": 0.0663,
+      "step": 766
+    },
+    {
+      "epoch": 4.29350104821803,
+      "grad_norm": 1.5551415143149132,
+      "learning_rate": 3.1114622257520004e-06,
+      "loss": 0.1021,
+      "step": 768
+    },
+    {
+      "epoch": 4.304682040531097,
+      "grad_norm": 1.836559018121584,
+      "learning_rate": 3.1029017604102655e-06,
+      "loss": 0.099,
+      "step": 770
+    },
+    {
+      "epoch": 4.315863032844165,
+      "grad_norm": 1.0818921388079483,
+      "learning_rate": 3.0943337828911673e-06,
+      "loss": 0.0899,
+      "step": 772
+    },
+    {
+      "epoch": 4.327044025157233,
+      "grad_norm": 0.9784785751112162,
+      "learning_rate": 3.085758399952011e-06,
+      "loss": 0.1016,
+      "step": 774
+    },
+    {
+      "epoch": 4.3382250174703,
+      "grad_norm": 1.348338975607883,
+      "learning_rate": 3.0771757184423716e-06,
+      "loss": 0.1063,
+      "step": 776
+    },
+    {
+      "epoch": 4.349406009783368,
+      "grad_norm": 2.1529902019434455,
+      "learning_rate": 3.0685858453027668e-06,
+      "loss": 0.089,
+      "step": 778
+    },
+    {
+      "epoch": 4.360587002096436,
+      "grad_norm": 1.3031273077449874,
+      "learning_rate": 3.0599888875633192e-06,
+      "loss": 0.1077,
+      "step": 780
+    },
+    {
+      "epoch": 4.371767994409504,
+      "grad_norm": 1.3772043306307704,
+      "learning_rate": 3.0513849523424298e-06,
+      "loss": 0.0879,
+      "step": 782
+    },
+    {
+      "epoch": 4.382948986722572,
+      "grad_norm": 1.7829225937512299,
+      "learning_rate": 3.0427741468454375e-06,
+      "loss": 0.1099,
+      "step": 784
+    },
+    {
+      "epoch": 4.39412997903564,
+      "grad_norm": 1.1143653742483424,
+      "learning_rate": 3.034156578363284e-06,
+      "loss": 0.0908,
+      "step": 786
+    },
+    {
+      "epoch": 4.405310971348707,
+      "grad_norm": 1.9841896768408593,
+      "learning_rate": 3.0255323542711784e-06,
+      "loss": 0.0846,
+      "step": 788
+    },
+    {
+      "epoch": 4.416491963661775,
+      "grad_norm": 1.1622503242476587,
+      "learning_rate": 3.0169015820272595e-06,
+      "loss": 0.0809,
+      "step": 790
+    },
+    {
+      "epoch": 4.427672955974843,
+      "grad_norm": 1.4138977756081776,
+      "learning_rate": 3.0082643691712572e-06,
+      "loss": 0.0832,
+      "step": 792
+    },
+    {
+      "epoch": 4.43885394828791,
+      "grad_norm": 1.3694425414816003,
+      "learning_rate": 2.9996208233231506e-06,
+      "loss": 0.1015,
+      "step": 794
+    },
+    {
+      "epoch": 4.450034940600978,
+      "grad_norm": 1.8252502558409327,
+      "learning_rate": 2.9909710521818265e-06,
+      "loss": 0.1049,
+      "step": 796
+    },
+    {
+      "epoch": 4.461215932914046,
+      "grad_norm": 1.4396307405101365,
+      "learning_rate": 2.9823151635237424e-06,
+      "loss": 0.0613,
+      "step": 798
+    },
+    {
+      "epoch": 4.472396925227114,
+      "grad_norm": 1.3667673153541864,
+      "learning_rate": 2.973653265201578e-06,
+      "loss": 0.1081,
+      "step": 800
+    },
+    {
+      "epoch": 4.483577917540182,
+      "grad_norm": 1.761976942384573,
+      "learning_rate": 2.964985465142895e-06,
+      "loss": 0.1002,
+      "step": 802
+    },
+    {
+      "epoch": 4.49475890985325,
+      "grad_norm": 1.6343471974417978,
+      "learning_rate": 2.9563118713487895e-06,
+      "loss": 0.0749,
+      "step": 804
+    },
+    {
+      "epoch": 4.505939902166317,
+      "grad_norm": 2.0454570442431046,
+      "learning_rate": 2.9476325918925484e-06,
+      "loss": 0.0857,
+      "step": 806
+    },
+    {
+      "epoch": 4.517120894479385,
+      "grad_norm": 1.7007295640066746,
+      "learning_rate": 2.938947734918302e-06,
+      "loss": 0.1085,
+      "step": 808
+    },
+    {
+      "epoch": 4.528301886792453,
+      "grad_norm": 1.5611422829954795,
+      "learning_rate": 2.9302574086396774e-06,
+      "loss": 0.0775,
+      "step": 810
+    },
+    {
+      "epoch": 4.53948287910552,
+      "grad_norm": 1.7913016893140525,
+      "learning_rate": 2.9215617213384494e-06,
+      "loss": 0.0875,
+      "step": 812
+    },
+    {
+      "epoch": 4.550663871418588,
+      "grad_norm": 1.5753063947599002,
+      "learning_rate": 2.91286078136319e-06,
+      "loss": 0.0805,
+      "step": 814
+    },
+    {
+      "epoch": 4.561844863731656,
+      "grad_norm": 1.8942921897754963,
+      "learning_rate": 2.904154697127921e-06,
+      "loss": 0.0806,
+      "step": 816
+    },
+    {
+      "epoch": 4.573025856044724,
+      "grad_norm": 1.791394910046461,
+      "learning_rate": 2.8954435771107604e-06,
+      "loss": 0.0992,
+      "step": 818
+    },
+    {
+      "epoch": 4.584206848357792,
+      "grad_norm": 1.245790765054016,
+      "learning_rate": 2.8867275298525743e-06,
+      "loss": 0.0886,
+      "step": 820
+    },
+    {
+      "epoch": 4.59538784067086,
+      "grad_norm": 1.5133863011334676,
+      "learning_rate": 2.878006663955621e-06,
+      "loss": 0.0886,
+      "step": 822
+    },
+    {
+      "epoch": 4.606568832983927,
+      "grad_norm": 2.0502622868705993,
+      "learning_rate": 2.8692810880821997e-06,
+      "loss": 0.0716,
+      "step": 824
+    },
+    {
+      "epoch": 4.617749825296995,
+      "grad_norm": 1.2876873289352964,
+      "learning_rate": 2.860550910953296e-06,
+      "loss": 0.0943,
+      "step": 826
+    },
+    {
+      "epoch": 4.628930817610063,
+      "grad_norm": 1.440475980645125,
+      "learning_rate": 2.8518162413472266e-06,
+      "loss": 0.1083,
+      "step": 828
+    },
+    {
+      "epoch": 4.64011180992313,
+      "grad_norm": 1.3754262878787067,
+      "learning_rate": 2.843077188098286e-06,
+      "loss": 0.1041,
+      "step": 830
+    },
+    {
+      "epoch": 4.651292802236198,
+      "grad_norm": 1.4424213259038674,
+      "learning_rate": 2.834333860095388e-06,
+      "loss": 0.0807,
+      "step": 832
+    },
+    {
+      "epoch": 4.662473794549266,
+      "grad_norm": 1.994638545215632,
+      "learning_rate": 2.8255863662807097e-06,
+      "loss": 0.0819,
+      "step": 834
+    },
+    {
+      "epoch": 4.673654786862334,
+      "grad_norm": 1.5478645240921063,
+      "learning_rate": 2.8168348156483356e-06,
+      "loss": 0.113,
+      "step": 836
+    },
+    {
+      "epoch": 4.684835779175402,
+      "grad_norm": 1.324879005941319,
+      "learning_rate": 2.8124575531000226e-06,
+      "loss": 0.11,
+      "step": 838
+    },
+    {
+      "epoch": 4.69601677148847,
+      "grad_norm": 1.5993247352100177,
+      "learning_rate": 2.803700121715214e-06,
+      "loss": 0.0903,
+      "step": 840
+    },
+    {
+      "epoch": 4.707197763801537,
+      "grad_norm": 1.256541482417978,
+      "learning_rate": 2.7949389062160946e-06,
+      "loss": 0.0925,
+      "step": 842
+    },
+    {
+      "epoch": 4.718378756114605,
+      "grad_norm": 2.706891920194882,
+      "learning_rate": 2.786174015767721e-06,
+      "loss": 0.084,
+      "step": 844
+    },
+    {
+      "epoch": 4.729559748427673,
+      "grad_norm": 1.3220515828132557,
+      "learning_rate": 2.7774055595809395e-06,
+      "loss": 0.0801,
+      "step": 846
+    },
+    {
+      "epoch": 4.7407407407407405,
+      "grad_norm": 1.5911477732332153,
+      "learning_rate": 2.768633646911027e-06,
+      "loss": 0.0938,
+      "step": 848
+    },
+    {
+      "epoch": 4.751921733053808,
+      "grad_norm": 1.1333988378482527,
+      "learning_rate": 2.759858387056325e-06,
+      "loss": 0.0721,
+      "step": 850
+    },
+    {
+      "epoch": 4.763102725366876,
+      "grad_norm": 1.4690260920140663,
+      "learning_rate": 2.7510798893568846e-06,
+      "loss": 0.0769,
+      "step": 852
+    },
+    {
+      "epoch": 4.774283717679944,
+      "grad_norm": 1.3785131166774844,
+      "learning_rate": 2.742298263193099e-06,
+      "loss": 0.1064,
+      "step": 854
+    },
+    {
+      "epoch": 4.785464709993012,
+      "grad_norm": 1.39128795327872,
+      "learning_rate": 2.733513617984342e-06,
+      "loss": 0.075,
+      "step": 856
+    },
+    {
+      "epoch": 4.79664570230608,
+      "grad_norm": 1.6826021403482612,
+      "learning_rate": 2.724726063187605e-06,
+      "loss": 0.1175,
+      "step": 858
+    },
+    {
+      "epoch": 4.8078266946191475,
+      "grad_norm": 1.353741266830404,
+      "learning_rate": 2.715935708296134e-06,
+      "loss": 0.1146,
+      "step": 860
+    },
+    {
+      "epoch": 4.819007686932215,
+      "grad_norm": 1.4488179633464906,
+      "learning_rate": 2.707142662838062e-06,
+      "loss": 0.1033,
+      "step": 862
+    },
+    {
+      "epoch": 4.830188679245283,
+      "grad_norm": 1.307354977462126,
+      "learning_rate": 2.6983470363750497e-06,
+      "loss": 0.093,
+      "step": 864
+    },
+    {
+      "epoch": 4.8413696715583505,
+      "grad_norm": 1.4753004858703918,
+      "learning_rate": 2.689548938500914e-06,
+      "loss": 0.0905,
+      "step": 866
+    },
+    {
+      "epoch": 4.852550663871418,
+      "grad_norm": 1.551558439927485,
+      "learning_rate": 2.6807484788402676e-06,
+      "loss": 0.075,
+      "step": 868
+    },
+    {
+      "epoch": 4.863731656184486,
+      "grad_norm": 1.499892261020302,
+      "learning_rate": 2.67194576704715e-06,
+      "loss": 0.0876,
+      "step": 870
+    },
+    {
+      "epoch": 4.8749126484975545,
+      "grad_norm": 1.82643381640813,
+      "learning_rate": 2.6631409128036637e-06,
+      "loss": 0.0892,
+      "step": 872
+    },
+    {
+      "epoch": 4.886093640810622,
+      "grad_norm": 1.3480606493487655,
+      "learning_rate": 2.6543340258186063e-06,
+      "loss": 0.0816,
+      "step": 874
+    },
+    {
+      "epoch": 4.89727463312369,
+      "grad_norm": 2.2307067144092407,
+      "learning_rate": 2.6455252158261015e-06,
+      "loss": 0.0994,
+      "step": 876
+    },
+    {
+      "epoch": 4.9084556254367575,
+      "grad_norm": 1.8646868858712458,
+      "learning_rate": 2.636714592584235e-06,
+      "loss": 0.0902,
+      "step": 878
+    },
+    {
+      "epoch": 4.919636617749825,
+      "grad_norm": 1.535171207325978,
+      "learning_rate": 2.6279022658736856e-06,
+      "loss": 0.0911,
+      "step": 880
+    },
+    {
+      "epoch": 4.930817610062893,
+      "grad_norm": 1.1594360070916991,
+      "learning_rate": 2.619088345496358e-06,
+      "loss": 0.066,
+      "step": 882
+    },
+    {
+      "epoch": 4.941998602375961,
+      "grad_norm": 1.6526631394475477,
+      "learning_rate": 2.610272941274012e-06,
+      "loss": 0.1014,
+      "step": 884
+    },
+    {
+      "epoch": 4.953179594689029,
+      "grad_norm": 1.8240816325874138,
+      "learning_rate": 2.6014561630468993e-06,
+      "loss": 0.0928,
+      "step": 886
+    },
+    {
+      "epoch": 4.964360587002097,
+      "grad_norm": 1.3816438884334348,
+      "learning_rate": 2.5926381206723885e-06,
+      "loss": 0.088,
+      "step": 888
+    },
+    {
+      "epoch": 4.9755415793151645,
+      "grad_norm": 1.3157397283692482,
+      "learning_rate": 2.583818924023601e-06,
+      "loss": 0.0938,
+      "step": 890
+    },
+    {
+      "epoch": 4.986722571628232,
+      "grad_norm": 1.464557516575305,
+      "learning_rate": 2.5749986829880423e-06,
+      "loss": 0.0781,
+      "step": 892
+    },
+    {
+      "epoch": 4.9979035639413,
+      "grad_norm": 1.8481309973872981,
+      "learning_rate": 2.5661775074662276e-06,
+      "loss": 0.0708,
+      "step": 894
+    },
+    {
+      "epoch": 5.0090845562543675,
+      "grad_norm": 1.3777408578534927,
+      "learning_rate": 2.5573555073703172e-06,
+      "loss": 0.0574,
+      "step": 896
+    },
+    {
+      "epoch": 5.020265548567435,
+      "grad_norm": 1.5585565063610693,
+      "learning_rate": 2.5485327926227464e-06,
+      "loss": 0.0533,
+      "step": 898
+    },
+    {
+      "epoch": 5.031446540880503,
+      "grad_norm": 3.8488829032344403,
+      "learning_rate": 2.539709473154855e-06,
+      "loss": 0.0524,
+      "step": 900
+    },
+    {
+      "epoch": 5.042627533193571,
+      "grad_norm": 1.360678519326562,
+      "learning_rate": 2.5308856589055164e-06,
+      "loss": 0.0608,
+      "step": 902
+    },
+    {
+      "epoch": 5.053808525506638,
+      "grad_norm": 1.4720850175627471,
+      "learning_rate": 2.5220614598197708e-06,
+      "loss": 0.0527,
+      "step": 904
+    },
+    {
+      "epoch": 5.064989517819707,
+      "grad_norm": 1.2412662972591795,
+      "learning_rate": 2.513236985847451e-06,
+      "loss": 0.0488,
+      "step": 906
+    },
+    {
+      "epoch": 5.0761705101327745,
+      "grad_norm": 1.3236580966844242,
+      "learning_rate": 2.5044123469418174e-06,
+      "loss": 0.0638,
+      "step": 908
+    },
+    {
+      "epoch": 5.087351502445842,
+      "grad_norm": 1.8348241342651854,
+      "learning_rate": 2.495587653058184e-06,
+      "loss": 0.0629,
+      "step": 910
+    },
+    {
+      "epoch": 5.09853249475891,
+      "grad_norm": 0.9662213920921242,
+      "learning_rate": 2.4867630141525493e-06,
+      "loss": 0.0722,
+      "step": 912
+    },
+    {
+      "epoch": 5.109713487071978,
+      "grad_norm": 1.6784486385619315,
+      "learning_rate": 2.477938540180231e-06,
+      "loss": 0.0482,
+      "step": 914
+    },
+    {
+      "epoch": 5.120894479385045,
+      "grad_norm": 1.386742744607905,
+      "learning_rate": 2.4691143410944844e-06,
+      "loss": 0.0596,
+      "step": 916
+    },
+    {
+      "epoch": 5.132075471698113,
+      "grad_norm": 1.5375835898995094,
+      "learning_rate": 2.4602905268451455e-06,
+      "loss": 0.0592,
+      "step": 918
+    },
+    {
+      "epoch": 5.143256464011181,
+      "grad_norm": 1.334707574114043,
+      "learning_rate": 2.451467207377254e-06,
+      "loss": 0.0493,
+      "step": 920
+    },
+    {
+      "epoch": 5.154437456324249,
+      "grad_norm": 1.018606004126685,
+      "learning_rate": 2.442644492629683e-06,
+      "loss": 0.0544,
+      "step": 922
+    },
+    {
+      "epoch": 5.165618448637317,
+      "grad_norm": 1.0236510244569192,
+      "learning_rate": 2.433822492533774e-06,
+      "loss": 0.0501,
+      "step": 924
+    },
+    {
+      "epoch": 5.176799440950385,
+      "grad_norm": 0.8191759766926784,
+      "learning_rate": 2.4250013170119585e-06,
+      "loss": 0.0594,
+      "step": 926
+    },
+    {
+      "epoch": 5.187980433263452,
+      "grad_norm": 1.0938612787512558,
+      "learning_rate": 2.4161810759763993e-06,
+      "loss": 0.0544,
+      "step": 928
+    },
+    {
+      "epoch": 5.19916142557652,
+      "grad_norm": 1.3602285379082586,
+      "learning_rate": 2.407361879327612e-06,
+      "loss": 0.0442,
+      "step": 930
+    },
+    {
+      "epoch": 5.210342417889588,
+      "grad_norm": 1.1380441045618945,
+      "learning_rate": 2.398543836953101e-06,
+      "loss": 0.0563,
+      "step": 932
+    },
+    {
+      "epoch": 5.221523410202655,
+      "grad_norm": 1.1080478505241853,
+      "learning_rate": 2.389727058725989e-06,
+      "loss": 0.0515,
+      "step": 934
+    },
+    {
+      "epoch": 5.232704402515723,
+      "grad_norm": 1.2558697950305333,
+      "learning_rate": 2.380911654503643e-06,
+      "loss": 0.0507,
+      "step": 936
+    },
+    {
+      "epoch": 5.243885394828791,
+      "grad_norm": 1.2293644348010904,
+      "learning_rate": 2.3720977341263152e-06,
+      "loss": 0.0607,
+      "step": 938
+    },
+    {
+      "epoch": 5.255066387141859,
+      "grad_norm": 1.292488994918762,
+      "learning_rate": 2.3632854074157653e-06,
+      "loss": 0.0474,
+      "step": 940
+    },
+    {
+      "epoch": 5.266247379454927,
+      "grad_norm": 1.2671492916227067,
+      "learning_rate": 2.3544747841738998e-06,
+      "loss": 0.0769,
+      "step": 942
+    },
+    {
+      "epoch": 5.277428371767995,
+      "grad_norm": 1.6102887076835615,
+      "learning_rate": 2.3456659741813945e-06,
+      "loss": 0.0496,
+      "step": 944
+    },
+    {
+      "epoch": 5.288609364081062,
+      "grad_norm": 1.577997048333656,
+      "learning_rate": 2.3368590871963367e-06,
+      "loss": 0.0796,
+      "step": 946
+    },
+    {
+      "epoch": 5.29979035639413,
+      "grad_norm": 2.278441135480121,
+      "learning_rate": 2.328054232952851e-06,
+      "loss": 0.0679,
+      "step": 948
+    },
+    {
+      "epoch": 5.310971348707198,
+      "grad_norm": 1.1443796744340577,
+      "learning_rate": 2.3192515211597332e-06,
+      "loss": 0.0589,
+      "step": 950
+    },
+    {
+      "epoch": 5.322152341020265,
+      "grad_norm": 1.3246252050774938,
+      "learning_rate": 2.3104510614990875e-06,
+      "loss": 0.0711,
+      "step": 952
+    },
+    {
+      "epoch": 5.333333333333333,
+      "grad_norm": 2.3404125762291574,
+      "learning_rate": 2.301652963624951e-06,
+      "loss": 0.0571,
+      "step": 954
+    },
+    {
+      "epoch": 5.344514325646401,
+      "grad_norm": 1.6173224098499974,
+      "learning_rate": 2.292857337161938e-06,
+      "loss": 0.0715,
+      "step": 956
+    },
+    {
+      "epoch": 5.355695317959469,
+      "grad_norm": 1.416375080557459,
+      "learning_rate": 2.2840642917038666e-06,
+      "loss": 0.0555,
+      "step": 958
+    },
+    {
+      "epoch": 5.366876310272537,
+      "grad_norm": 1.2819320119071211,
+      "learning_rate": 2.2752739368123948e-06,
+      "loss": 0.0486,
+      "step": 960
+    },
+    {
+      "epoch": 5.378057302585605,
+      "grad_norm": 1.1198977788924485,
+      "learning_rate": 2.2664863820156593e-06,
+      "loss": 0.0408,
+      "step": 962
+    },
+    {
+      "epoch": 5.389238294898672,
+      "grad_norm": 1.1451798114445098,
+      "learning_rate": 2.2577017368069017e-06,
+      "loss": 0.0626,
+      "step": 964
+    },
+    {
+      "epoch": 5.40041928721174,
+      "grad_norm": 1.3380127274735694,
+      "learning_rate": 2.248920110643116e-06,
+      "loss": 0.0568,
+      "step": 966
+    },
+    {
+      "epoch": 5.411600279524808,
+      "grad_norm": 1.4489239240672898,
+      "learning_rate": 2.2401416129436753e-06,
+      "loss": 0.059,
+      "step": 968
+    },
+    {
+      "epoch": 5.422781271837875,
+      "grad_norm": 1.3130908635170957,
+      "learning_rate": 2.2313663530889734e-06,
+      "loss": 0.0444,
+      "step": 970
+    },
+    {
+      "epoch": 5.433962264150943,
+      "grad_norm": 1.2045728193533076,
+      "learning_rate": 2.222594440419061e-06,
+      "loss": 0.0952,
+      "step": 972
+    },
+    {
+      "epoch": 5.445143256464011,
+      "grad_norm": 1.1505612686257871,
+      "learning_rate": 2.2138259842322794e-06,
+      "loss": 0.0536,
+      "step": 974
+    },
+    {
+      "epoch": 5.456324248777079,
+      "grad_norm": 1.521719008832957,
+      "learning_rate": 2.2050610937839058e-06,
+      "loss": 0.073,
+      "step": 976
+    },
+    {
+      "epoch": 5.467505241090147,
+      "grad_norm": 1.3381824532405695,
+      "learning_rate": 2.1962998782847863e-06,
+      "loss": 0.0583,
+      "step": 978
+    },
+    {
+      "epoch": 5.478686233403215,
+      "grad_norm": 1.1782879600371732,
+      "learning_rate": 2.1875424468999787e-06,
+      "loss": 0.052,
+      "step": 980
+    },
+    {
+      "epoch": 5.489867225716282,
+      "grad_norm": 1.1689516819440322,
+      "learning_rate": 2.178788908747387e-06,
+      "loss": 0.0515,
+      "step": 982
+    },
+    {
+      "epoch": 5.50104821802935,
+      "grad_norm": 1.1479989981730907,
+      "learning_rate": 2.170039372896409e-06,
+      "loss": 0.055,
+      "step": 984
+    },
+    {
+      "epoch": 5.512229210342418,
+      "grad_norm": 1.3922562574409854,
+      "learning_rate": 2.161293948366573e-06,
+      "loss": 0.0554,
+      "step": 986
+    },
+    {
+      "epoch": 5.523410202655485,
+      "grad_norm": 1.409490849880991,
+      "learning_rate": 2.152552744126178e-06,
+      "loss": 0.0392,
+      "step": 988
+    },
+    {
+      "epoch": 5.534591194968553,
+      "grad_norm": 1.2479629003574995,
+      "learning_rate": 2.1438158690909413e-06,
+      "loss": 0.0599,
+      "step": 990
+    },
+    {
+      "epoch": 5.545772187281621,
+      "grad_norm": 1.2371376050465024,
+      "learning_rate": 2.1350834321226344e-06,
+      "loss": 0.0664,
+      "step": 992
+    },
+    {
+      "epoch": 5.556953179594689,
+      "grad_norm": 1.593505278104288,
+      "learning_rate": 2.126355542027734e-06,
+      "loss": 0.0479,
+      "step": 994
+    },
+    {
+      "epoch": 5.568134171907757,
+      "grad_norm": 1.2742537988695015,
+      "learning_rate": 2.117632307556059e-06,
+      "loss": 0.0803,
+      "step": 996
+    },
+    {
+      "epoch": 5.579315164220825,
+      "grad_norm": 1.3748039610126324,
+      "learning_rate": 2.1089138373994226e-06,
+      "loss": 0.0416,
+      "step": 998
+    },
+    {
+      "epoch": 5.590496156533892,
+      "grad_norm": 2.4084571636039755,
+      "learning_rate": 2.100200240190273e-06,
+      "loss": 0.0514,
+      "step": 1000
+    },
+    {
+      "epoch": 5.60167714884696,
+      "grad_norm": 1.1933752040503858,
+      "learning_rate": 2.09149162450034e-06,
+      "loss": 0.0625,
+      "step": 1002
+    },
+    {
+      "epoch": 5.612858141160028,
+      "grad_norm": 1.037709039674537,
+      "learning_rate": 2.0827880988392856e-06,
+      "loss": 0.0514,
+      "step": 1004
+    },
+    {
+      "epoch": 5.6240391334730955,
+      "grad_norm": 1.315142680072312,
+      "learning_rate": 2.0740897716533475e-06,
+      "loss": 0.0593,
+      "step": 1006
+    },
+    {
+      "epoch": 5.635220125786163,
+      "grad_norm": 1.0531660230737552,
+      "learning_rate": 2.0653967513239934e-06,
+      "loss": 0.0543,
+      "step": 1008
+    },
+    {
+      "epoch": 5.646401118099231,
+      "grad_norm": 1.2633776013551097,
+      "learning_rate": 2.0567091461665636e-06,
+      "loss": 0.0431,
+      "step": 1010
+    },
+    {
+      "epoch": 5.657582110412299,
+      "grad_norm": 1.449959564050197,
+      "learning_rate": 2.0480270644289282e-06,
+      "loss": 0.0482,
+      "step": 1012
+    },
+    {
+      "epoch": 5.668763102725367,
+      "grad_norm": 1.1071912059302882,
+      "learning_rate": 2.0393506142901347e-06,
+      "loss": 0.0564,
+      "step": 1014
+    },
+    {
+      "epoch": 5.679944095038435,
+      "grad_norm": 0.9876137346535111,
+      "learning_rate": 2.0306799038590595e-06,
+      "loss": 0.0391,
+      "step": 1016
+    },
+    {
+      "epoch": 5.6911250873515025,
+      "grad_norm": 1.1071464038310999,
+      "learning_rate": 2.0220150411730638e-06,
+      "loss": 0.0636,
+      "step": 1018
+    },
+    {
+      "epoch": 5.70230607966457,
+      "grad_norm": 1.0473491285671832,
+      "learning_rate": 2.013356134196643e-06,
+      "loss": 0.0581,
+      "step": 1020
+    },
+    {
+      "epoch": 5.713487071977638,
+      "grad_norm": 1.1296902267336801,
+      "learning_rate": 2.004703290820086e-06,
+      "loss": 0.0604,
+      "step": 1022
+    },
+    {
+      "epoch": 5.7246680642907055,
+      "grad_norm": 1.309317661735025,
+      "learning_rate": 1.9960566188581306e-06,
+      "loss": 0.0438,
+      "step": 1024
+    },
+    {
+      "epoch": 5.735849056603773,
+      "grad_norm": 0.8918766336417149,
+      "learning_rate": 1.9874162260486146e-06,
+      "loss": 0.0475,
+      "step": 1026
+    },
+    {
+      "epoch": 5.747030048916841,
+      "grad_norm": 1.2095534019736167,
+      "learning_rate": 1.978782220051142e-06,
+      "loss": 0.0454,
+      "step": 1028
+    },
+    {
+      "epoch": 5.7582110412299095,
+      "grad_norm": 1.1967009451687045,
+      "learning_rate": 1.9701547084457314e-06,
+      "loss": 0.0697,
+      "step": 1030
+    },
+    {
+      "epoch": 5.769392033542977,
+      "grad_norm": 1.8160556667087309,
+      "learning_rate": 1.961533798731486e-06,
+      "loss": 0.0422,
+      "step": 1032
+    },
+    {
+      "epoch": 5.780573025856045,
+      "grad_norm": 1.590627053883797,
+      "learning_rate": 1.952919598325247e-06,
+      "loss": 0.0602,
+      "step": 1034
+    },
+    {
+      "epoch": 5.7917540181691125,
+      "grad_norm": 1.4584761134724722,
+      "learning_rate": 1.944312214560256e-06,
+      "loss": 0.0575,
+      "step": 1036
+    },
+    {
+      "epoch": 5.80293501048218,
+      "grad_norm": 1.6093909025543798,
+      "learning_rate": 1.935711754684824e-06,
+      "loss": 0.0814,
+      "step": 1038
+    },
+    {
+      "epoch": 5.814116002795248,
+      "grad_norm": 1.7715253484509736,
+      "learning_rate": 1.9271183258609836e-06,
+      "loss": 0.0608,
+      "step": 1040
+    },
+    {
+      "epoch": 5.825296995108316,
+      "grad_norm": 0.850327251905485,
+      "learning_rate": 1.9185320351631654e-06,
+      "loss": 0.0388,
+      "step": 1042
+    },
+    {
+      "epoch": 5.836477987421384,
+      "grad_norm": 1.4837292387797913,
+      "learning_rate": 1.9099529895768552e-06,
+      "loss": 0.0567,
+      "step": 1044
+    },
+    {
+      "epoch": 5.847658979734452,
+      "grad_norm": 1.0384213631474088,
+      "learning_rate": 1.901381295997267e-06,
+      "loss": 0.0661,
+      "step": 1046
+    },
+    {
+      "epoch": 5.8588399720475195,
+      "grad_norm": 1.2071171218984706,
+      "learning_rate": 1.8928170612280067e-06,
+      "loss": 0.0665,
+      "step": 1048
+    },
+    {
+      "epoch": 5.870020964360587,
+      "grad_norm": 1.2020194163974407,
+      "learning_rate": 1.8842603919797436e-06,
+      "loss": 0.0466,
+      "step": 1050
+    },
+    {
+      "epoch": 5.881201956673655,
+      "grad_norm": 1.141150946131999,
+      "learning_rate": 1.8757113948688827e-06,
+      "loss": 0.0562,
+      "step": 1052
+    },
+    {
+      "epoch": 5.8923829489867225,
+      "grad_norm": 1.583487458549684,
+      "learning_rate": 1.8671701764162287e-06,
+      "loss": 0.0589,
+      "step": 1054
+    },
+    {
+      "epoch": 5.90356394129979,
+      "grad_norm": 1.3417276690702418,
+      "learning_rate": 1.8586368430456708e-06,
+      "loss": 0.0604,
+      "step": 1056
+    },
+    {
+      "epoch": 5.914744933612858,
+      "grad_norm": 1.3294273305641617,
+      "learning_rate": 1.8501115010828423e-06,
+      "loss": 0.0628,
+      "step": 1058
+    },
+    {
+      "epoch": 5.925925925925926,
+      "grad_norm": 1.2448945324282268,
+      "learning_rate": 1.8415942567538106e-06,
+      "loss": 0.0554,
+      "step": 1060
+    },
+    {
+      "epoch": 5.937106918238994,
+      "grad_norm": 0.960687093766239,
+      "learning_rate": 1.8330852161837399e-06,
+      "loss": 0.0532,
+      "step": 1062
+    },
+    {
+      "epoch": 5.948287910552062,
+      "grad_norm": 1.4656893110825278,
+      "learning_rate": 1.8245844853955786e-06,
+      "loss": 0.0719,
+      "step": 1064
+    },
+    {
+      "epoch": 5.9594689028651295,
+      "grad_norm": 1.6634277575338297,
+      "learning_rate": 1.8160921703087368e-06,
+      "loss": 0.0565,
+      "step": 1066
+    },
+    {
+      "epoch": 5.970649895178197,
+      "grad_norm": 1.7257111050609335,
+      "learning_rate": 1.8076083767377595e-06,
+      "loss": 0.068,
+      "step": 1068
+    },
+    {
+      "epoch": 5.981830887491265,
+      "grad_norm": 1.42483183153276,
+      "learning_rate": 1.7991332103910184e-06,
+      "loss": 0.0613,
+      "step": 1070
+    },
+    {
+      "epoch": 5.993011879804333,
+      "grad_norm": 1.4316025881020678,
+      "learning_rate": 1.7906667768693853e-06,
+      "loss": 0.0481,
+      "step": 1072
+    },
+    {
+      "epoch": 6.0041928721174,
+      "grad_norm": 1.037376667784287,
+      "learning_rate": 1.782209181664924e-06,
+      "loss": 0.0483,
+      "step": 1074
+    },
+    {
+      "epoch": 6.015373864430468,
+      "grad_norm": 1.0336168566598631,
+      "learning_rate": 1.773760530159571e-06,
+      "loss": 0.0347,
+      "step": 1076
+    },
+    {
+      "epoch": 6.026554856743536,
+      "grad_norm": 0.7872905184564322,
+      "learning_rate": 1.7653209276238242e-06,
+      "loss": 0.0355,
+      "step": 1078
+    },
+    {
+      "epoch": 6.037735849056604,
+      "grad_norm": 1.772389302776251,
+      "learning_rate": 1.7568904792154328e-06,
+      "loss": 0.0542,
+      "step": 1080
+    },
+    {
+      "epoch": 6.048916841369672,
+      "grad_norm": 1.3577848873845724,
+      "learning_rate": 1.7484692899780812e-06,
+      "loss": 0.0583,
+      "step": 1082
+    },
+    {
+      "epoch": 6.06009783368274,
+      "grad_norm": 0.7840766650439943,
+      "learning_rate": 1.740057464840088e-06,
+      "loss": 0.0289,
+      "step": 1084
+    },
+    {
+      "epoch": 6.071278825995807,
+      "grad_norm": 0.9255675051401594,
+      "learning_rate": 1.7316551086130925e-06,
+      "loss": 0.0417,
+      "step": 1086
+    },
+    {
+      "epoch": 6.082459818308875,
+      "grad_norm": 0.9107219582827843,
+      "learning_rate": 1.7232623259907538e-06,
+      "loss": 0.0429,
+      "step": 1088
+    },
+    {
+      "epoch": 6.093640810621943,
+      "grad_norm": 1.0296310110561282,
+      "learning_rate": 1.714879221547439e-06,
+      "loss": 0.0362,
+      "step": 1090
+    },
+    {
+      "epoch": 6.10482180293501,
+      "grad_norm": 0.9575340239366315,
+      "learning_rate": 1.7065058997369288e-06,
+      "loss": 0.0471,
+      "step": 1092
+    },
+    {
+      "epoch": 6.116002795248078,
+      "grad_norm": 0.7430183397758778,
+      "learning_rate": 1.6981424648911112e-06,
+      "loss": 0.0351,
+      "step": 1094
+    },
+    {
+      "epoch": 6.127183787561146,
+      "grad_norm": 0.9807593854080312,
+      "learning_rate": 1.6897890212186804e-06,
+      "loss": 0.0334,
+      "step": 1096
+    },
+    {
+      "epoch": 6.138364779874214,
+      "grad_norm": 1.2961448011313597,
+      "learning_rate": 1.6814456728038431e-06,
+      "loss": 0.025,
+      "step": 1098
+    },
+    {
+      "epoch": 6.149545772187282,
+      "grad_norm": 0.961636779671174,
+      "learning_rate": 1.673112523605015e-06,
+      "loss": 0.0285,
+      "step": 1100
+    },
+    {
+      "epoch": 6.16072676450035,
+      "grad_norm": 0.9647606646620928,
+      "learning_rate": 1.6647896774535324e-06,
+      "loss": 0.0303,
+      "step": 1102
+    },
+    {
+      "epoch": 6.171907756813417,
+      "grad_norm": 1.1381988477100318,
+      "learning_rate": 1.6564772380523546e-06,
+      "loss": 0.0358,
+      "step": 1104
+    },
+    {
+      "epoch": 6.183088749126485,
+      "grad_norm": 0.7901346245952422,
+      "learning_rate": 1.648175308974771e-06,
+      "loss": 0.0279,
+      "step": 1106
+    },
+    {
+      "epoch": 6.194269741439553,
+      "grad_norm": 1.2717247572933381,
+      "learning_rate": 1.6398839936631142e-06,
+      "loss": 0.0328,
+      "step": 1108
+    },
+    {
+      "epoch": 6.20545073375262,
+      "grad_norm": 1.2916496315117834,
+      "learning_rate": 1.631603395427466e-06,
+      "loss": 0.055,
+      "step": 1110
+    },
+    {
+      "epoch": 6.216631726065688,
+      "grad_norm": 0.9740099844597652,
+      "learning_rate": 1.6233336174443762e-06,
+      "loss": 0.048,
+      "step": 1112
+    },
+    {
+      "epoch": 6.227812718378756,
+      "grad_norm": 1.0103830292004847,
+      "learning_rate": 1.6150747627555713e-06,
+      "loss": 0.0434,
+      "step": 1114
+    },
+    {
+      "epoch": 6.238993710691824,
+      "grad_norm": 1.1350854047223082,
+      "learning_rate": 1.6068269342666749e-06,
+      "loss": 0.0389,
+      "step": 1116
+    },
+    {
+      "epoch": 6.250174703004892,
+      "grad_norm": 0.7884154494279628,
+      "learning_rate": 1.5985902347459239e-06,
+      "loss": 0.0432,
+      "step": 1118
+    },
+    {
+      "epoch": 6.26135569531796,
+      "grad_norm": 0.8788178903528164,
+      "learning_rate": 1.5903647668228855e-06,
+      "loss": 0.0432,
+      "step": 1120
+    },
+    {
+      "epoch": 6.272536687631027,
+      "grad_norm": 0.6393918351108393,
+      "learning_rate": 1.5821506329871834e-06,
+      "loss": 0.0253,
+      "step": 1122
+    },
+    {
+      "epoch": 6.283717679944095,
+      "grad_norm": 1.0870268262489273,
+      "learning_rate": 1.5739479355872162e-06,
+      "loss": 0.0364,
+      "step": 1124
+    },
+    {
+      "epoch": 6.294898672257163,
+      "grad_norm": 1.1679875063936556,
+      "learning_rate": 1.5657567768288868e-06,
+      "loss": 0.0333,
+      "step": 1126
+    },
+    {
+      "epoch": 6.30607966457023,
+      "grad_norm": 0.8388447320245327,
+      "learning_rate": 1.5575772587743222e-06,
+      "loss": 0.0316,
+      "step": 1128
+    },
+    {
+      "epoch": 6.317260656883298,
+      "grad_norm": 0.7710273725047172,
+      "learning_rate": 1.5494094833406092e-06,
+      "loss": 0.0308,
+      "step": 1130
+    },
+    {
+      "epoch": 6.328441649196366,
+      "grad_norm": 1.3107972415612894,
+      "learning_rate": 1.5412535522985205e-06,
+      "loss": 0.0186,
+      "step": 1132
+    },
+    {
+      "epoch": 6.339622641509434,
+      "grad_norm": 0.8488196487806184,
+      "learning_rate": 1.5331095672712463e-06,
+      "loss": 0.023,
+      "step": 1134
+    },
+    {
+      "epoch": 6.350803633822502,
+      "grad_norm": 1.014050814471419,
+      "learning_rate": 1.5249776297331302e-06,
+      "loss": 0.0425,
+      "step": 1136
+    },
+    {
+      "epoch": 6.36198462613557,
+      "grad_norm": 0.8160528908459946,
+      "learning_rate": 1.516857841008401e-06,
+      "loss": 0.0407,
+      "step": 1138
+    },
+    {
+      "epoch": 6.373165618448637,
+      "grad_norm": 0.6924190623075557,
+      "learning_rate": 1.5087503022699168e-06,
+      "loss": 0.0527,
+      "step": 1140
+    },
+    {
+      "epoch": 6.384346610761705,
+      "grad_norm": 1.0149043689805195,
+      "learning_rate": 1.5006551145378967e-06,
+      "loss": 0.0367,
+      "step": 1142
+    },
+    {
+      "epoch": 6.395527603074773,
+      "grad_norm": 1.5920991707794845,
+      "learning_rate": 1.4925723786786691e-06,
+      "loss": 0.0319,
+      "step": 1144
+    },
+    {
+      "epoch": 6.40670859538784,
+      "grad_norm": 0.8834798218634231,
+      "learning_rate": 1.4845021954034106e-06,
+      "loss": 0.0372,
+      "step": 1146
+    },
+    {
+      "epoch": 6.417889587700908,
+      "grad_norm": 1.072104658850445,
+      "learning_rate": 1.476444665266889e-06,
+      "loss": 0.0413,
+      "step": 1148
+    },
+    {
+      "epoch": 6.429070580013976,
+      "grad_norm": 1.1893734124292998,
+      "learning_rate": 1.4683998886662187e-06,
+      "loss": 0.0307,
+      "step": 1150
+    },
+    {
+      "epoch": 6.440251572327044,
+      "grad_norm": 1.1513167005422524,
+      "learning_rate": 1.4603679658396006e-06,
+      "loss": 0.0402,
+      "step": 1152
+    },
+    {
+      "epoch": 6.451432564640112,
+      "grad_norm": 1.0586602700365229,
+      "learning_rate": 1.4523489968650795e-06,
+      "loss": 0.0303,
+      "step": 1154
+    },
+    {
+      "epoch": 6.46261355695318,
+      "grad_norm": 0.7650987855999634,
+      "learning_rate": 1.4443430816592936e-06,
+      "loss": 0.0312,
+      "step": 1156
+    },
+    {
+      "epoch": 6.473794549266247,
+      "grad_norm": 0.7470083708652993,
+      "learning_rate": 1.4363503199762296e-06,
+      "loss": 0.0298,
+      "step": 1158
+    },
+    {
+      "epoch": 6.484975541579315,
+      "grad_norm": 1.2247183517462086,
+      "learning_rate": 1.4283708114059853e-06,
+      "loss": 0.0476,
+      "step": 1160
+    },
+    {
+      "epoch": 6.496156533892383,
+      "grad_norm": 1.0042001049340177,
+      "learning_rate": 1.4204046553735174e-06,
+      "loss": 0.0421,
+      "step": 1162
+    },
+    {
+      "epoch": 6.5073375262054505,
+      "grad_norm": 1.0066856707214424,
+      "learning_rate": 1.4124519511374158e-06,
+      "loss": 0.0277,
+      "step": 1164
+    },
+    {
+      "epoch": 6.518518518518518,
+      "grad_norm": 1.3761888161849996,
+      "learning_rate": 1.404512797788657e-06,
+      "loss": 0.0251,
+      "step": 1166
+    },
+    {
+      "epoch": 6.529699510831586,
+      "grad_norm": 0.7445041473181229,
+      "learning_rate": 1.396587294249374e-06,
+      "loss": 0.0383,
+      "step": 1168
+    },
+    {
+      "epoch": 6.540880503144654,
+      "grad_norm": 1.0231799225570892,
+      "learning_rate": 1.3886755392716225e-06,
+      "loss": 0.0289,
+      "step": 1170
+    },
+    {
+      "epoch": 6.552061495457722,
+      "grad_norm": 1.0842064444530823,
+      "learning_rate": 1.3807776314361498e-06,
+      "loss": 0.0341,
+      "step": 1172
+    },
+    {
+      "epoch": 6.56324248777079,
+      "grad_norm": 0.9409388421938562,
+      "learning_rate": 1.3728936691511704e-06,
+      "loss": 0.0413,
+      "step": 1174
+    },
+    {
+      "epoch": 6.5744234800838575,
+      "grad_norm": 0.8052329748698783,
+      "learning_rate": 1.3650237506511333e-06,
+      "loss": 0.0399,
+      "step": 1176
+    },
+    {
+      "epoch": 6.585604472396925,
+      "grad_norm": 0.6879172446908371,
+      "learning_rate": 1.3571679739955029e-06,
+      "loss": 0.0288,
+      "step": 1178
+    },
+    {
+      "epoch": 6.596785464709993,
+      "grad_norm": 0.8737080494275846,
+      "learning_rate": 1.3493264370675352e-06,
+      "loss": 0.0181,
+      "step": 1180
+    },
+    {
+      "epoch": 6.6079664570230605,
+      "grad_norm": 0.8744184416405667,
+      "learning_rate": 1.3414992375730587e-06,
+      "loss": 0.0432,
+      "step": 1182
+    },
+    {
+      "epoch": 6.619147449336128,
+      "grad_norm": 0.9265074156931595,
+      "learning_rate": 1.3336864730392587e-06,
+      "loss": 0.0464,
+      "step": 1184
+    },
+    {
+      "epoch": 6.630328441649196,
+      "grad_norm": 1.14003149718633,
+      "learning_rate": 1.3258882408134582e-06,
+      "loss": 0.0271,
+      "step": 1186
+    },
+    {
+      "epoch": 6.6415094339622645,
+      "grad_norm": 0.8949105583359471,
+      "learning_rate": 1.3181046380619078e-06,
+      "loss": 0.0276,
+      "step": 1188
+    },
+    {
+      "epoch": 6.652690426275332,
+      "grad_norm": 1.0602768370905677,
+      "learning_rate": 1.3103357617685746e-06,
+      "loss": 0.0352,
+      "step": 1190
+    },
+    {
+      "epoch": 6.6638714185884,
+      "grad_norm": 1.187406942024327,
+      "learning_rate": 1.3025817087339335e-06,
+      "loss": 0.0597,
+      "step": 1192
+    },
+    {
+      "epoch": 6.6750524109014675,
+      "grad_norm": 0.8451020033143687,
+      "learning_rate": 1.2948425755737592e-06,
+      "loss": 0.0359,
+      "step": 1194
+    },
+    {
+      "epoch": 6.686233403214535,
+      "grad_norm": 1.2760921925255864,
+      "learning_rate": 1.2871184587179286e-06,
+      "loss": 0.0285,
+      "step": 1196
+    },
+    {
+      "epoch": 6.697414395527603,
+      "grad_norm": 0.7781748766075295,
+      "learning_rate": 1.2794094544092111e-06,
+      "loss": 0.0346,
+      "step": 1198
+    },
+    {
+      "epoch": 6.7085953878406706,
+      "grad_norm": 1.1832623077309767,
+      "learning_rate": 1.2717156587020746e-06,
+      "loss": 0.041,
+      "step": 1200
+    },
+    {
+      "epoch": 6.719776380153739,
+      "grad_norm": 1.3133094357866473,
+      "learning_rate": 1.2640371674614866e-06,
+      "loss": 0.0629,
+      "step": 1202
+    },
+    {
+      "epoch": 6.730957372466807,
+      "grad_norm": 0.7218331862903847,
+      "learning_rate": 1.2563740763617198e-06,
+      "loss": 0.0366,
+      "step": 1204
+    },
+    {
+      "epoch": 6.7421383647798745,
+      "grad_norm": 0.9560652150388108,
+      "learning_rate": 1.2487264808851654e-06,
+      "loss": 0.044,
+      "step": 1206
+    },
+    {
+      "epoch": 6.753319357092942,
+      "grad_norm": 1.1190106870390395,
+      "learning_rate": 1.2410944763211302e-06,
+      "loss": 0.0517,
+      "step": 1208
+    },
+    {
+      "epoch": 6.76450034940601,
+      "grad_norm": 0.7835985914687663,
+      "learning_rate": 1.2334781577646653e-06,
+      "loss": 0.0272,
+      "step": 1210
+    },
+    {
+      "epoch": 6.7756813417190775,
+      "grad_norm": 2.056446636497986,
+      "learning_rate": 1.2258776201153702e-06,
+      "loss": 0.0239,
+      "step": 1212
+    },
+    {
+      "epoch": 6.786862334032145,
+      "grad_norm": 0.8485551422736736,
+      "learning_rate": 1.218292958076213e-06,
+      "loss": 0.0206,
+      "step": 1214
+    },
+    {
+      "epoch": 6.798043326345213,
+      "grad_norm": 1.2531964534501892,
+      "learning_rate": 1.2107242661523544e-06,
+      "loss": 0.0254,
+      "step": 1216
+    },
+    {
+      "epoch": 6.809224318658281,
+      "grad_norm": 1.269537638790587,
+      "learning_rate": 1.203171638649962e-06,
+      "loss": 0.0299,
+      "step": 1218
+    },
+    {
+      "epoch": 6.820405310971349,
+      "grad_norm": 1.1178764385402225,
+      "learning_rate": 1.195635169675045e-06,
+      "loss": 0.0396,
+      "step": 1220
+    },
+    {
+      "epoch": 6.831586303284417,
+      "grad_norm": 0.6920818283019613,
+      "learning_rate": 1.1881149531322744e-06,
+      "loss": 0.0268,
+      "step": 1222
+    },
+    {
+      "epoch": 6.8427672955974845,
+      "grad_norm": 0.80369354175751,
+      "learning_rate": 1.180611082723814e-06,
+      "loss": 0.031,
+      "step": 1224
+    },
+    {
+      "epoch": 6.853948287910552,
+      "grad_norm": 0.7447389756775401,
+      "learning_rate": 1.1731236519481593e-06,
+      "loss": 0.0345,
+      "step": 1226
+    },
+    {
+      "epoch": 6.86512928022362,
+      "grad_norm": 1.1115305000722167,
+      "learning_rate": 1.1656527540989595e-06,
+      "loss": 0.0283,
+      "step": 1228
+    },
+    {
+      "epoch": 6.876310272536688,
+      "grad_norm": 1.2279572164110593,
+      "learning_rate": 1.1581984822638706e-06,
+      "loss": 0.0452,
+      "step": 1230
+    },
+    {
+      "epoch": 6.887491264849755,
+      "grad_norm": 0.8467749629186313,
+      "learning_rate": 1.1507609293233837e-06,
+      "loss": 0.0283,
+      "step": 1232
+    },
+    {
+      "epoch": 6.898672257162823,
+      "grad_norm": 1.355703618365484,
+      "learning_rate": 1.1433401879496723e-06,
+      "loss": 0.0366,
+      "step": 1234
+    },
+    {
+      "epoch": 6.909853249475891,
+      "grad_norm": 1.004917827499692,
+      "learning_rate": 1.135936350605438e-06,
+      "loss": 0.0496,
+      "step": 1236
+    },
+    {
+      "epoch": 6.921034241788959,
+      "grad_norm": 1.2615070307313305,
+      "learning_rate": 1.1285495095427563e-06,
+      "loss": 0.0461,
+      "step": 1238
+    },
+    {
+      "epoch": 6.932215234102027,
+      "grad_norm": 0.9861185460727813,
+      "learning_rate": 1.1211797568019312e-06,
+      "loss": 0.0366,
+      "step": 1240
+    },
+    {
+      "epoch": 6.943396226415095,
+      "grad_norm": 1.6576290169923233,
+      "learning_rate": 1.113827184210343e-06,
+      "loss": 0.0337,
+      "step": 1242
+    },
+    {
+      "epoch": 6.954577218728162,
+      "grad_norm": 1.1363579065284033,
+      "learning_rate": 1.1064918833813073e-06,
+      "loss": 0.0406,
+      "step": 1244
+    },
+    {
+      "epoch": 6.96575821104123,
+      "grad_norm": 1.3125191134965577,
+      "learning_rate": 1.0991739457129333e-06,
+      "loss": 0.0397,
+      "step": 1246
+    },
+    {
+      "epoch": 6.976939203354298,
+      "grad_norm": 0.8904462468667067,
+      "learning_rate": 1.0918734623869835e-06,
+      "loss": 0.0407,
+      "step": 1248
+    },
+    {
+      "epoch": 6.988120195667365,
+      "grad_norm": 2.263233580582389,
+      "learning_rate": 1.0845905243677416e-06,
+      "loss": 0.0307,
+      "step": 1250
+    },
+    {
+      "epoch": 6.999301187980433,
+      "grad_norm": 0.791294534235276,
+      "learning_rate": 1.0773252224008726e-06,
+      "loss": 0.0387,
+      "step": 1252
+    },
+    {
+      "epoch": 7.010482180293501,
+      "grad_norm": 0.76599595030522,
+      "learning_rate": 1.0700776470122981e-06,
+      "loss": 0.0269,
+      "step": 1254
+    },
+    {
+      "epoch": 7.021663172606569,
+      "grad_norm": 0.7331796337642835,
+      "learning_rate": 1.0628478885070647e-06,
+      "loss": 0.0221,
+      "step": 1256
+    },
+    {
+      "epoch": 7.032844164919637,
+      "grad_norm": 0.6845784469587074,
+      "learning_rate": 1.05563603696822e-06,
+      "loss": 0.0291,
+      "step": 1258
+    },
+    {
+      "epoch": 7.044025157232705,
+      "grad_norm": 0.8176233505690059,
+      "learning_rate": 1.0484421822556904e-06,
+      "loss": 0.0364,
+      "step": 1260
+    },
+    {
+      "epoch": 7.055206149545772,
+      "grad_norm": 0.8629657573128657,
+      "learning_rate": 1.041266414005162e-06,
+      "loss": 0.0265,
+      "step": 1262
+    },
+    {
+      "epoch": 7.06638714185884,
+      "grad_norm": 1.1172499462707595,
+      "learning_rate": 1.0341088216269625e-06,
+      "loss": 0.0157,
+      "step": 1264
+    },
+    {
+      "epoch": 7.077568134171908,
+      "grad_norm": 0.5230775744769823,
+      "learning_rate": 1.0269694943049462e-06,
+      "loss": 0.0157,
+      "step": 1266
+    },
+    {
+      "epoch": 7.088749126484975,
+      "grad_norm": 0.8978199171663125,
+      "learning_rate": 1.0198485209953865e-06,
+      "loss": 0.0275,
+      "step": 1268
+    },
+    {
+      "epoch": 7.099930118798043,
+      "grad_norm": 0.815308309594077,
+      "learning_rate": 1.0127459904258621e-06,
+      "loss": 0.0237,
+      "step": 1270
+    },
+    {
+      "epoch": 7.111111111111111,
+      "grad_norm": 0.8967571058386815,
+      "learning_rate": 1.0056619910941592e-06,
+      "loss": 0.019,
+      "step": 1272
+    },
+    {
+      "epoch": 7.122292103424179,
+      "grad_norm": 0.7843358442700527,
+      "learning_rate": 9.98596611267158e-07,
+      "loss": 0.021,
+      "step": 1274
+    },
+    {
+      "epoch": 7.133473095737247,
+      "grad_norm": 0.6797830063456453,
+      "learning_rate": 9.915499389797444e-07,
+      "loss": 0.0316,
+      "step": 1276
+    },
+    {
+      "epoch": 7.144654088050315,
+      "grad_norm": 0.6688875199025872,
+      "learning_rate": 9.845220620337054e-07,
+      "loss": 0.0303,
+      "step": 1278
+    },
+    {
+      "epoch": 7.155835080363382,
+      "grad_norm": 0.6664970872749731,
+      "learning_rate": 9.77513067996636e-07,
+      "loss": 0.0219,
+      "step": 1280
+    },
+    {
+      "epoch": 7.16701607267645,
+      "grad_norm": 0.7973098520727987,
+      "learning_rate": 9.705230442008542e-07,
+      "loss": 0.0376,
+      "step": 1282
+    },
+    {
+      "epoch": 7.178197064989518,
+      "grad_norm": 0.8759703504057706,
+      "learning_rate": 9.63552077742301e-07,
+      "loss": 0.0385,
+      "step": 1284
+    },
+    {
+      "epoch": 7.189378057302585,
+      "grad_norm": 1.0267904937054426,
+      "learning_rate": 9.56600255479469e-07,
+      "loss": 0.0222,
+      "step": 1286
+    },
+    {
+      "epoch": 7.200559049615653,
+      "grad_norm": 0.6389768145894307,
+      "learning_rate": 9.4966766403231e-07,
+      "loss": 0.018,
+      "step": 1288
+    },
+    {
+      "epoch": 7.211740041928721,
+      "grad_norm": 0.5762313893158477,
+      "learning_rate": 9.427543897811584e-07,
+      "loss": 0.0165,
+      "step": 1290
+    },
+    {
+      "epoch": 7.222921034241789,
+      "grad_norm": 0.5902518126138557,
+      "learning_rate": 9.358605188656603e-07,
+      "loss": 0.02,
+      "step": 1292
+    },
+    {
+      "epoch": 7.234102026554857,
+      "grad_norm": 0.824105561963567,
+      "learning_rate": 9.289861371836886e-07,
+      "loss": 0.0337,
+      "step": 1294
+    },
+    {
+      "epoch": 7.245283018867925,
+      "grad_norm": 0.504698332550927,
+      "learning_rate": 9.22131330390286e-07,
+      "loss": 0.0283,
+      "step": 1296
+    },
+    {
+      "epoch": 7.256464011180992,
+      "grad_norm": 0.5789695393721453,
+      "learning_rate": 9.152961838965879e-07,
+      "loss": 0.0169,
+      "step": 1298
+    },
+    {
+      "epoch": 7.26764500349406,
+      "grad_norm": 1.4892687104014115,
+      "learning_rate": 9.084807828687628e-07,
+      "loss": 0.0314,
+      "step": 1300
+    },
+    {
+      "epoch": 7.278825995807128,
+      "grad_norm": 1.0727067281323632,
+      "learning_rate": 9.016852122269493e-07,
+      "loss": 0.0274,
+      "step": 1302
+    },
+    {
+      "epoch": 7.290006988120195,
+      "grad_norm": 0.7309629553367788,
+      "learning_rate": 8.949095566441985e-07,
+      "loss": 0.0219,
+      "step": 1304
+    },
+    {
+      "epoch": 7.301187980433263,
+      "grad_norm": 0.6871990809680889,
+      "learning_rate": 8.881539005454215e-07,
+      "loss": 0.0339,
+      "step": 1306
+    },
+    {
+      "epoch": 7.312368972746331,
+      "grad_norm": 0.8530617423198913,
+      "learning_rate": 8.814183281063326e-07,
+      "loss": 0.0248,
+      "step": 1308
+    },
+    {
+      "epoch": 7.323549965059399,
+      "grad_norm": 0.76651991997128,
+      "learning_rate": 8.747029232524037e-07,
+      "loss": 0.023,
+      "step": 1310
+    },
+    {
+      "epoch": 7.334730957372467,
+      "grad_norm": 0.6966547986519114,
+      "learning_rate": 8.680077696578182e-07,
+      "loss": 0.0332,
+      "step": 1312
+    },
+    {
+      "epoch": 7.345911949685535,
+      "grad_norm": 1.0873098335521205,
+      "learning_rate": 8.613329507444274e-07,
+      "loss": 0.0234,
+      "step": 1314
+    },
+    {
+      "epoch": 7.357092941998602,
+      "grad_norm": 0.6461932986017782,
+      "learning_rate": 8.546785496807116e-07,
+      "loss": 0.0242,
+      "step": 1316
+    },
+    {
+      "epoch": 7.36827393431167,
+      "grad_norm": 0.7614414460885182,
+      "learning_rate": 8.480446493807464e-07,
+      "loss": 0.031,
+      "step": 1318
+    },
+    {
+      "epoch": 7.379454926624738,
+      "grad_norm": 0.641294466328584,
+      "learning_rate": 8.414313325031642e-07,
+      "loss": 0.028,
+      "step": 1320
+    },
+    {
+      "epoch": 7.3906359189378055,
+      "grad_norm": 0.47088954187562415,
+      "learning_rate": 8.348386814501286e-07,
+      "loss": 0.0186,
+      "step": 1322
+    },
+    {
+      "epoch": 7.401816911250873,
+      "grad_norm": 0.7909087034714356,
+      "learning_rate": 8.282667783663056e-07,
+      "loss": 0.0212,
+      "step": 1324
+    },
+    {
+      "epoch": 7.412997903563941,
+      "grad_norm": 0.8059238279425677,
+      "learning_rate": 8.217157051378411e-07,
+      "loss": 0.0239,
+      "step": 1326
+    },
+    {
+      "epoch": 7.424178895877009,
+      "grad_norm": 0.788531385863816,
+      "learning_rate": 8.151855433913414e-07,
+      "loss": 0.0199,
+      "step": 1328
+    },
+    {
+      "epoch": 7.435359888190077,
+      "grad_norm": 1.1393964476120448,
+      "learning_rate": 8.086763744928536e-07,
+      "loss": 0.0292,
+      "step": 1330
+    },
+    {
+      "epoch": 7.446540880503145,
+      "grad_norm": 0.5408108502649198,
+      "learning_rate": 8.02188279546853e-07,
+      "loss": 0.0146,
+      "step": 1332
+    },
+    {
+      "epoch": 7.4577218728162125,
+      "grad_norm": 0.8749206113652656,
+      "learning_rate": 7.957213393952335e-07,
+      "loss": 0.0247,
+      "step": 1334
+    },
+    {
+      "epoch": 7.46890286512928,
+      "grad_norm": 0.7053824386402378,
+      "learning_rate": 7.892756346162986e-07,
+      "loss": 0.02,
+      "step": 1336
+    },
+    {
+      "epoch": 7.480083857442348,
+      "grad_norm": 0.6965900833846856,
+      "learning_rate": 7.82851245523761e-07,
+      "loss": 0.0315,
+      "step": 1338
+    },
+    {
+      "epoch": 7.4912648497554155,
+      "grad_norm": 0.9392067120327887,
+      "learning_rate": 7.764482521657343e-07,
+      "loss": 0.0308,
+      "step": 1340
+    },
+    {
+      "epoch": 7.502445842068483,
+      "grad_norm": 0.7074561491918046,
+      "learning_rate": 7.700667343237453e-07,
+      "loss": 0.0171,
+      "step": 1342
+    },
+    {
+      "epoch": 7.513626834381551,
+      "grad_norm": 0.7697005768650605,
+      "learning_rate": 7.637067715117327e-07,
+      "loss": 0.0302,
+      "step": 1344
+    },
+    {
+      "epoch": 7.5248078266946195,
+      "grad_norm": 1.176668146060272,
+      "learning_rate": 7.573684429750583e-07,
+      "loss": 0.0265,
+      "step": 1346
+    },
+    {
+      "epoch": 7.535988819007687,
+      "grad_norm": 0.7258573280389607,
+      "learning_rate": 7.510518276895234e-07,
+      "loss": 0.0257,
+      "step": 1348
+    },
+    {
+      "epoch": 7.547169811320755,
+      "grad_norm": 1.1195611459347754,
+      "learning_rate": 7.447570043603755e-07,
+      "loss": 0.0261,
+      "step": 1350
+    },
+    {
+      "epoch": 7.5583508036338225,
+      "grad_norm": 0.9527258409378455,
+      "learning_rate": 7.384840514213404e-07,
+      "loss": 0.0524,
+      "step": 1352
+    },
+    {
+      "epoch": 7.56953179594689,
+      "grad_norm": 0.7074898357644916,
+      "learning_rate": 7.322330470336314e-07,
+      "loss": 0.0205,
+      "step": 1354
+    },
+    {
+      "epoch": 7.580712788259958,
+      "grad_norm": 0.9361424266631929,
+      "learning_rate": 7.26004069084987e-07,
+      "loss": 0.0217,
+      "step": 1356
+    },
+    {
+      "epoch": 7.5918937805730256,
+      "grad_norm": 1.7048958108176762,
+      "learning_rate": 7.197971951886956e-07,
+      "loss": 0.0225,
+      "step": 1358
+    },
+    {
+      "epoch": 7.603074772886094,
+      "grad_norm": 0.8812767707258257,
+      "learning_rate": 7.13612502682623e-07,
+      "loss": 0.0196,
+      "step": 1360
+    },
+    {
+      "epoch": 7.614255765199162,
+      "grad_norm": 0.5682027618905875,
+      "learning_rate": 7.074500686282609e-07,
+      "loss": 0.019,
+      "step": 1362
+    },
+    {
+      "epoch": 7.6254367575122295,
+      "grad_norm": 0.4475598932931596,
+      "learning_rate": 7.013099698097539e-07,
+      "loss": 0.0171,
+      "step": 1364
+    },
+    {
+      "epoch": 7.636617749825297,
+      "grad_norm": 0.5527498039813922,
+      "learning_rate": 6.951922827329535e-07,
+      "loss": 0.0217,
+      "step": 1366
+    },
+    {
+      "epoch": 7.647798742138365,
+      "grad_norm": 0.7984442985333638,
+      "learning_rate": 6.890970836244574e-07,
+      "loss": 0.0361,
+      "step": 1368
+    },
+    {
+      "epoch": 7.6589797344514325,
+      "grad_norm": 0.624268450810696,
+      "learning_rate": 6.830244484306623e-07,
+      "loss": 0.0158,
+      "step": 1370
+    },
+    {
+      "epoch": 7.6701607267645,
+      "grad_norm": 0.7493822409267487,
+      "learning_rate": 6.769744528168207e-07,
+      "loss": 0.0286,
+      "step": 1372
+    },
+    {
+      "epoch": 7.681341719077568,
+      "grad_norm": 0.6787647092695418,
+      "learning_rate": 6.709471721660904e-07,
+      "loss": 0.0215,
+      "step": 1374
+    },
+    {
+      "epoch": 7.692522711390636,
+      "grad_norm": 0.7321502006735149,
+      "learning_rate": 6.649426815786045e-07,
+      "loss": 0.0311,
+      "step": 1376
+    },
+    {
+      "epoch": 7.703703703703704,
+      "grad_norm": 0.701610396870259,
+      "learning_rate": 6.589610558705284e-07,
+      "loss": 0.0235,
+      "step": 1378
+    },
+    {
+      "epoch": 7.714884696016772,
+      "grad_norm": 0.6530846520546149,
+      "learning_rate": 6.53002369573131e-07,
+      "loss": 0.0245,
+      "step": 1380
+    },
+    {
+      "epoch": 7.7260656883298395,
+      "grad_norm": 0.7531427984254183,
+      "learning_rate": 6.470666969318554e-07,
+      "loss": 0.0315,
+      "step": 1382
+    },
+    {
+      "epoch": 7.737246680642907,
+      "grad_norm": 0.7301669272251805,
+      "learning_rate": 6.41154111905393e-07,
+      "loss": 0.0225,
+      "step": 1384
+    },
+    {
+      "epoch": 7.748427672955975,
+      "grad_norm": 0.8707140120777088,
+      "learning_rate": 6.352646881647647e-07,
+      "loss": 0.0259,
+      "step": 1386
+    },
+    {
+      "epoch": 7.759608665269043,
+      "grad_norm": 0.837200588883093,
+      "learning_rate": 6.29398499092399e-07,
+      "loss": 0.0474,
+      "step": 1388
+    },
+    {
+      "epoch": 7.77078965758211,
+      "grad_norm": 0.973530488120086,
+      "learning_rate": 6.235556177812205e-07,
+      "loss": 0.0329,
+      "step": 1390
+    },
+    {
+      "epoch": 7.781970649895178,
+      "grad_norm": 0.5813627298678434,
+      "learning_rate": 6.177361170337376e-07,
+      "loss": 0.0194,
+      "step": 1392
+    },
+    {
+      "epoch": 7.793151642208246,
+      "grad_norm": 0.8597088367336019,
+      "learning_rate": 6.119400693611358e-07,
+      "loss": 0.0123,
+      "step": 1394
+    },
+    {
+      "epoch": 7.804332634521314,
+      "grad_norm": 0.8368570476462492,
+      "learning_rate": 6.061675469823763e-07,
+      "loss": 0.0227,
+      "step": 1396
+    },
+    {
+      "epoch": 7.815513626834382,
+      "grad_norm": 0.5203392914919558,
+      "learning_rate": 6.004186218232933e-07,
+      "loss": 0.0217,
+      "step": 1398
+    },
+    {
+      "epoch": 7.82669461914745,
+      "grad_norm": 0.8572153440435842,
+      "learning_rate": 5.946933655156976e-07,
+      "loss": 0.0294,
+      "step": 1400
+    },
+    {
+      "epoch": 7.837875611460517,
+      "grad_norm": 0.6862577628733875,
+      "learning_rate": 5.889918493964869e-07,
+      "loss": 0.0228,
+      "step": 1402
+    },
+    {
+      "epoch": 7.849056603773585,
+      "grad_norm": 0.7097594226614418,
+      "learning_rate": 5.833141445067541e-07,
+      "loss": 0.0113,
+      "step": 1404
+    },
+    {
+      "epoch": 7.860237596086653,
+      "grad_norm": 0.6322499286175502,
+      "learning_rate": 5.776603215909041e-07,
+      "loss": 0.0229,
+      "step": 1406
+    },
+    {
+      "epoch": 7.87141858839972,
+      "grad_norm": 0.6798739232739857,
+      "learning_rate": 5.720304510957722e-07,
+      "loss": 0.0257,
+      "step": 1408
+    },
+    {
+      "epoch": 7.882599580712788,
+      "grad_norm": 0.6568708401714163,
+      "learning_rate": 5.66424603169744e-07,
+      "loss": 0.0285,
+      "step": 1410
+    },
+    {
+      "epoch": 7.893780573025856,
+      "grad_norm": 1.1483908878505031,
+      "learning_rate": 5.608428476618843e-07,
+      "loss": 0.0235,
+      "step": 1412
+    },
+    {
+      "epoch": 7.904961565338924,
+      "grad_norm": 0.9297111790590921,
+      "learning_rate": 5.552852541210651e-07,
+      "loss": 0.022,
+      "step": 1414
+    },
+    {
+      "epoch": 7.916142557651992,
+      "grad_norm": 0.7288896652277049,
+      "learning_rate": 5.497518917950986e-07,
+      "loss": 0.033,
+      "step": 1416
+    },
+    {
+      "epoch": 7.92732354996506,
+      "grad_norm": 1.3241630685241197,
+      "learning_rate": 5.44242829629878e-07,
+      "loss": 0.0236,
+      "step": 1418
+    },
+    {
+      "epoch": 7.938504542278127,
+      "grad_norm": 0.6616696784338312,
+      "learning_rate": 5.387581362685112e-07,
+      "loss": 0.03,
+      "step": 1420
+    },
+    {
+      "epoch": 7.949685534591195,
+      "grad_norm": 0.9223806906428696,
+      "learning_rate": 5.332978800504742e-07,
+      "loss": 0.0234,
+      "step": 1422
+    },
+    {
+      "epoch": 7.960866526904263,
+      "grad_norm": 1.1302104401143789,
+      "learning_rate": 5.278621290107533e-07,
+      "loss": 0.0334,
+      "step": 1424
+    },
+    {
+      "epoch": 7.97204751921733,
+      "grad_norm": 0.6145924647383543,
+      "learning_rate": 5.224509508789987e-07,
+      "loss": 0.0205,
+      "step": 1426
+    },
+    {
+      "epoch": 7.983228511530398,
+      "grad_norm": 0.6724718918142113,
+      "learning_rate": 5.170644130786842e-07,
+      "loss": 0.0315,
+      "step": 1428
+    },
+    {
+      "epoch": 7.994409503843466,
+      "grad_norm": 0.5897709957691004,
+      "learning_rate": 5.117025827262598e-07,
+      "loss": 0.0189,
+      "step": 1430
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1780,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 598197676277760.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}